Professional Documents
Culture Documents
output bit sequence of a CA. This implies An LFSR can be implemented using only a
that the same linear dependencies exist in few XOR gates whereas a CA requires at
the output bit stream of a CA as in the least one XOR gate for each node. This fact
output of the similar LFSR. brings up an obvious draw-back of CA:
Higher area overhead involved in
LFSR and CA are characterized by their implementation of CA compared to an
transition matrices, the analysis of these LFSR. So, the designer has to pay a penalty
matrices along with simulations give the on the area over-head by choosing CA over
measure of the randomness in the patterns LFSR.
generated; these measures show the higher
randomness of patterns produced by CAs. In ACA the communication is generally
Parallel patterns generated by LFSRs (using local, being restricted to the nearest
outputs from different nodes of an LFSR) neighboring cells and cells are regular and
have a strong correlation between each other topologically equivalent to one another but
due to the shifting of data. Pattern in the case of LFSR these property does not
generation in CAs does not involve shifting exist [2].
of data.
Following table shows the summary of
There is greater probability of an error in comparison
LFSR by aliasing compared to ACA due to Characteristics LFSR ACA
shifting of data in LFSR [1]. In case of CA Performance Very good Good- no
each node value is a function of the in case of feedback
neighboring nodes resulting in a lower internal path and
probability of an error. The presence of feedback, maximum
XOR gates in the feedback path of an poor for
on XOR
External Feedback LFSR and lack of a external gate
feedback path in a null boundary condition feedback between
results in higher operating speed for CAs. node
Randomness of Low- High – no
LFSR have a feedback from their end nodes; generated shifting of shifting of
this means a redesign of the LFSR is needed pattern bit causes bit
if the pattern length has to be changed. This correlation
is not the case with ACA. ACA is logically between
connected to their only to their neighbors pattern
and there is no feed-back for an ACA CAD Friendly No- Yes- node
employing the null boundary condition [5]. requires can be
Therefore, the pattern length generated by redesign for cascaded
CAs can be easily changed by cascading the change in easily
nodes. The regular structure of the nodes for pattern
CA makes them ideal for CAD tools by length
providing the much needed flexibility in Speed Lower than Higher than
design. However, it is difficult to construct a CA LFSR
maximum length sequence CA as compared Error Greater- by Lower- no
to an LFSR which can be constructed using Probability aliasing due shifting of
the primitive polynomials which are very to shifting bit
well documented.
COMO1O1-3
CONFERENCE ON ―SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)‖ MARCH 26-27 2011
COMO1O1-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO1O2-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
We apply the proposed procedure to segment model prior. We will see details on HMM-based
aerial images with two classes, “man- made region” and segmentation in the next section.
“natural region”, which could be used for target recognition S2) we compute FPRs and TPRs of classifiers in C and plot
and tracking. Here, the cost of misclassifying targets (or them. This defines the region in the ROC space. Further,
man-made regions) is higher than that of misclassifying we get the ROC convex hull region of .
non targets (or natural regions) to targets. The results show S3) we find the tangential point between the line with slope
that, as the misclassification cost of man-made regions in (2) and the ROC convex hull region . The classifier
increases, the HMM based segmenter tends to classify both corresponding to the found tangential point is the most
man-made and natural regions to man-made ones. Thus, the cost-effective classifier.
rate of falsely classifying into man-made regions increases.
The remainder of this paper is organized as follows. We
introduce ROC convex analysis in section II. We study its
application to HMM based segmentation procedure in
section III. In section IV, We apply the proposed
procedure to segment an aerial image. Section V concludes
this paper.
COMO1O2-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
unknown parameters γ and β; the parameter γ is the between iso-cost lines and . The approximated boundary,
parameter that represents the magnitude of magnetization denoted by , is a curve from (0, 0) to (1, 1). Finally,
of the random field which implies the dominance of class P is approximated by the convex hull of , and the optimal
against class N in common words; it is also related to the classifier is found using the procedure in Section II.
ratio of the prior probability of P to N; β is the parameter to
measure the strength of spatial coherence. For example, the IV. EXPERIMENTAL RESULTS
HMGMM model [11] uses the generalized bond
percolation (BP) model. Let Z = {} with Z = – 1 or 1: Z = 1 In this section, we apply the procedure in Section
and Z = – 1 implies the class P and N, respectively. The III to segment the aerial image with HMGMM with
generalized BP assumes that the probability of Z is generalized BP model in (4). The aerial image is composed
(4) of many sub-blocks which are classified into “natural”
Where (Z) ( (Z)) is the number of concordant (dis- regions or “man-made” regions. We call the “natural” sub-
concordant) adjacent pairs which are neighbors to each block as “negative” and “man-made” one as “positive.”
other. Here, ) is the partition function that is The HMGMM model has two parameters γ and β which
reflects the overall portion of man-made and spatial
coherence between adjacent sub-blocks, respectively. Each
introduces a classifier, say , and a point of
(FPR, TPR) in the ROC space. We let be D the collection
of all these s. The input image taken is
illustrated below.
COMO1O2-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
VI.REFERENCES
COMO1O2-4
Performance Comparison of VANET Routing Protocols
Fig 2: Packet Delivery Ratio with various vehicle distances Fig 4: Throughput with various vehicle distances
Figure 1 shows the Jitter Rate variation at Figure 3 shows the Routing Overhead at various
different distances between vehicles. It is distances between vehicles. It is observed that
observed that DSR Jitter Rate remains lowest DSR Routing Overhead is lowest, AODV and
throughput the interval of observation, while DSDV Routing Overhead remain similar from
DSDV and AODV show comparatively higher 20m to 100m but after 100m AODV Routing
Jitter rate. Overhead is going down.
Figure 2 shows the Packet Delivery Ratio at Figure 4 shows the Throughput with various
various distances between vehicles. It is observed vehicle distances is shown. Clear inference from
that DSR Packet Delivery Ratio remains high graph DSR Throughput is highest. AODV
during entire duration of observation. AODV Throughput remains in between other two and
Packet Delivery Ratio in between other two and DSDV Throughput is lowest among all three.
DSDV Packet Delivery Ratio is lowest among
three.
1
scheduler and resources is assumed constant, also the jobs are Schd_valueij = Strt_timej + (Job_reqi/Proc_powerj)
assumed to submitted on layer 1 having different job If Min is greater than Schd_valueij
requirement. Then
An algorithm is proposed in this paper which is suitable for Min = Schd_valueij
static job submission in heterogeneous resource environment Res_id = Rj
connected to the scheduler through homogeneous End If
communication environment. Greedy approach is used to End For
solve the job scheduling problem. According to the greedy Submit the job Ji to Res_id resource
approach “A greedy algorithm always makes the choice that Submit the input data of Ji job to Res_id resource
looks best at that moment. That is, it makes a locally optimal End SUBMIT_NEW_JOB
choice in the hope that this choice will lead to a globally
optimal solution" [8]. The proposed algorithm uses the similar Once the scheduler submits a job to a resource, the resource
approach; it takes every job as independent of each other and will remains for some time in processing of that job. The
each of them is scheduled on a resource to give minimum UPDATE_STATUS algorithm is used to find out when the
turnaround time for that job.The overall turnaround time of all resource will be available to process a new job. The
the jobs is thus minimized. The parameters used in this UPDATE_STATUS algorithm is given below:
algorithm are as follows:
A set of resources, R = {R1, R2, R3,........, Rn}. UPDATE_STATUS
Ji = The submitted ith job. /* Res_id is the resource on which the job Ji is submitted. j
Arr_timei = Arrival time of job Ji. is the index of resource on which the job Ji is submitted
Proc_powerj = Processing power of resource Rj. and Rj = Res_id*/
Strt_timej = Estimated time at which a job starts execution at Strt_timej = Schd_valueij
resource Rj. End UPDATE_STATUS
Job_reqi = Length of job Ji.
Schd_valueij = Expected turnaround time of ith job at jth The above presented algorithm has the time complexity of
resource. O(n) for each job, where n is the number of resources. The
Min = The minimum of Schd_valueij among all resources. above algorithm required additional space to store the resorces
Res_id = Current selected resource id having optimum current status for availability.
turnaround time.
The algorithm used to schedule a job is given as follows: IV. EXPERIMENTAL RESULTS
The GridSim simulator [6] is used to simulate the algorithms.
GREEDY_SCHEDULE The GridSim toolkit is used to simulate heterogeneous
/*The users submit their jobs on the scheduler.*/ resource environment and the communication environment.
For all resource Rj The experiments are performed with three algorithms. The
/*Initialize the start time at resources.*/ algorithms are Random Resource Selection and Equal Job
Strt_timej=0.0 Distribution and Proposed Algorithm. The input data is taken
End For to be the same for all the three algorithms. The simulation is
/*The jobs are stored in a queue Q.*/ conducted with three resources which are shown in Table 1.
Insert all the jobs Ji in Q
While Q is not empty TABLE 1: Resources with their architecture and processing power.
do Resource R0 R1 R2
Delete the job Ji Architecture Sun Ultra Sun Ultra Sun Ultra
SUBMIT_NEW_JOB
OS Unix AIX Unix
UPDATE_STATUS
Proc_power(in MIPS) 48000 43000 54000
Advance the Q pointer
End While
End GREEDY_SCHEDULE The scheduler submits these jobs on resources according to
these algorithms. The algorithms are presented one by one
The scheduler uses SUBMIT_NEW_JOB algorithm to find with their simulation results.
the best suited resource that minimizes the turnaround time.
The turnaround time is calculated on the basis of expected A. Random Resource Selection
completion time of a job. The detailed SUBMIT_NEW_JOB In this algorithm the scheduler contacts GIS to obtain the
algorithm is as follows: resource information and then it chooses a resource randomly
[7]. The job is submitted on this chosen resource. This
SUBMIT_JOB algorithm is very simple to implement and has less overhead
Min = ∞ on the scheduler. The bar chart (Fig. 2) shows the turnaround
For every resource Rj time of different jobs. The completion time is a time at which
/* Calculating the expected turnaround time*/
2
the result of a job is available. After simulation the average C. Proposed Algorithm
turnaround time is found to be 20105.65 seconds and all the
In Proposed Algorithm, the scheduler finds the resource
jobs are completed at the 64420.25th second.
information with the help of GIS and calculates the
approximate completion time of this job on every resource.
Using these values the scheduler chooses a resource which has
the minimum of completion time and submits that job on this
resource. The turnaround time of each job is shown in bar
chart in Fig. 4. Through this algorithm the average turnaround
time of these jobs is 17208.77 seconds and all the jobs are
completed at 41840.88th second. The Proposed Algorithm
further reduces the average turnaround time by 4.22% as
compared with Equal Job Distribution. The completion time
of all jobs takes some more time than Equal Job Distribution
algorithm.
The turnaround time of each job is shown by the bar chart in The proposed scheduling algorithm reduces the average
Fig. 3. Experimental results show that the average turnaround turnaround time of all submitted jobs. The considered
time is 17968.55 seconds and the last result is outputed at environment executed the jobs on different resources which
39000.22th second. Equal Job Distribution reduces the average are geographically distributed. It is observed that the Proposed
turnaround time by 10.62% and it takes less time in Algorithm reduces the average turnaround by 4.22% with
comparison to the Random Resource Selection to give all the Equal Job Distribution (as shown in Table 2). The algorithm
results. uses meta-scheduler where resource failure is not considered.
3
Region (HPC Asia 2000), Beijing, China. IEEE Computer
Society Press, USA, 2000.
REFERENCES
[5] Cong Liu, Sanjeev Baskiyar and Shuang Li, “A General
Distributed Scalable Peer to Peer for Mixed Tasks in Grids",
[1] Ammar H. Alhusaini, Viktor K. Prasanna,C.S.
Raghavendra, “Unified Resource Scheduling Framework for High Performance Computing – HiPC 2007, ISBN:978-3-
540-77219-4, 320-330, 2007.
Heterogeneous Computing Environments", in Proceedings of
the Eighth Heterogeneous Computing Workshop, San Juan,
[6] Rajkumar Buyya, Manzur Murshed, “GridSim: a toolkit
Puerto Rico, pp. 156-165, 1999.
for the modeling and simulation of distributed resource
management and scheduling for Grid computing", Technical
[2] N. Muthuvelu, J. Liu, N. L. Soe, S.r Venugopal, A.
Sulistio and R. Buyya, “A Dynamic Job Grouping-Based Report, Monash University, Nov. 2001. To appear in the
Journal of Concurrency and Computation: Practice and
Scheduling for Deploying Applications with Fine-Grained
Tasks on Global Grids", Proceedings of the 3rd Australasian Experience (CCPE), pp. 1-32, Wiley Press, May 2002.
Workshop on Grid Computing and e-Research (AusGrid
2005), Newcastle, Australia, 41-48, January 30 - February 4, [7] Volker Hamscher, Uwe Schewiegelshohn, Achim Streit,
Ramin Yahyapour, “Evaluation of Job-Scheduling Strategies
2005.
for Grid Computing", in 1st IEEE/ACM International
Workshop on Grid Computing (Grid 2000), Berlin, Lecture
[3] I. Foster, C Kesselman, “The Grid: Blueprint for a new
computing infrastructure", Morgan Kaufmann Publishers, San Notes in Computer Science (LNCS), Springer, Berlin,
Heidelberg, New York, pp. 191-202, 2000.[8] Cormen TH,
Francisco, USA, 1999.
[4] R. Buyya, D. Abramson, J.Giddy, “Nimrod/G: An Leiserson CE, Rivest RL, “Introduction to algorithms 2nd
edition", MIT and McGraw-Hill Book Company, Boston
Architecture for a Resource Management and Scheduling
System in a Global Computation Grid", International Massachusetts, cp. 16, 370-403, 2001.
Conference on High Performance Computing in Asia-Pacific
4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
router did not interpret this field until it traffic and the best example of this kind of
gives any kind of service parameters. In this traffic is email[9] , [10].
TOS byte three fields were reserved for IP
precedence and other three for type of (iii) Low priority traffic : This kind of
service[12] , [13]. Higher the IP precedence traffic does not need any guarantee , priority
data would treated more priority and lower or significance and it is not required to
the IP precedence data would be dealt with deliver the traffic at its appropriate
low priority . destination . This kind of traffic includes the
codes generated by a hacker or spam kind of
During the initial stage of internet it was mails[18].
not needed in the network because
commercialization was not so much but as it MPLS:
grown up the need of different type of Multi Protocol Label Switching is a
service were also grown up[16]. new technology for getting the fast and rapid
transfer of data. High speed convergence of
Classification of traffic based on QOS
Classification of different flows of traffic scalability is possible in MPLS based
into classes may depend on several network [1],[2],[3]. Multi Protocol Label
parameters. Basically similar data packets Switching has become the first choice for
are considered in the same class[18]. The packet transportation , which fulfills the
most common way to classify the flows may several requirements of next generation[19].
depend on header fields such as IP Several service providers are going to
precedence and DSCP fields[15]. One of the deploy the MPLS on a common platform to
header of TCP is also used for classification achieve convergence of existing
of traffic by identifying the length of coming technologies like X.25 ; ATM / Frame relay
packets or by checking the MAC address of (FR) and best effort services. It provides the
both senders and receivers address[14]. transmission resources by providing the Diff
When the traffic is classified three main Serve to a network. MPLS provides the
classes comes out – intelligent routing and gives the important
improvements in the switching performance
(i) High priority for sensitive traffic : The of the network instead of what networks
traffic like real time requires some special architecture is. MPLS also provides
treatment like no delay and less jitter .This scalability like advantages for Virtual
comes in High Priority traffic and most Private Network (VPN) and maintain end-
common example of such traffic is to-end QOS [1] , [3] , [5].
VoIP[10].
In normal IP based networks IP
(ii) Best effort traffic : When traffic does routers performs destination based routing.
not need any delay related guarantees . In when they have to sent traffic router always
this case packets are send after high priority uses the simple shortest path first algorithm
to compute the shortest path between
COMO1O6-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO1O6-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO1O6-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Table 6. O/P Statics of MPLS Network The statistics collected from both the
networks and MPLS gives better result than
(1) Packet forwarding in MPLS is depends normal IP routed networks. Packet of data
on labels rather then IP prefixes , which faces longer delays in IP based networks and
reduces overhead of processing of routers. other factors were also improved in MPLS
(2) It has been observed that features of load based networks. All the necessary
balancing of MPLS removes the unwanted configurations of routers have been recorded
congestion from the network. and two different scenarios were compared.
Then routing tables of IP based networks
(3) Practical scenario proved that QOS and Cisco Express Forwarding (CEF) tables
works better in MPLS than IP routing. of MPLS networks have been obtained.
labels”, AT&T Labs–Research Shannon Published on May 1, 2003, v0.2 May 15,
Laboratory 180 Park Avenue Florham Park, 2003, v0.3 September 10, 2003
2003.
[13]. Traffic conditioning factors
[5]. Luc De Ghein, “MPLS Fundamentals”, http://www.cisco.com/en/US/tech/tk543/tk7
Cisco Systems, Cisco Press 800 East 96th 57/technologies_tech_note09186a00800949f
Street Indianapolis, ISBN: 1-58705-197-4, 2.shtml http://en.wikipedia.org/ wiki/
2007. IP_traceback# Packet_marking,
http://en.wikipedia.org/wiki/Traffic_policing
[6]. Benjamin Tang, Ahmet A. Akyamac, ,http://en.wikipedia.org/wiki/Traffic_shapin
Chi-Hung Kelvin Chu and Ramesh g# Traffic_Classification(Access Date
Nagarajan, “MPLS Network Requirements January 12, 2009)
and Design for Carriers: Wireline and
Wireless Case Studies”, [14]. Carter Horney for Nuntius Systems,
Telecommunications Network strategy and Quality of Service and MPLS (White
planning symposium, pages 1-6, Nov 2006. Paper), Inc.13700 Alton Pkwy., Suite 154-
266 Irvine, CA 92612 949.295.0475 voice
[7]. Muzammil Ahmad Khan, “Quantitative www.nuntius.com
Analysis of Multiprotocol Label Switching
(MPLS)”, Student Confrence, Volume 1, [15]. P. Prabagaran & Joseph
Issue 16-17, page(s): 56-65, Aug 2002. B,“Experiences with Class of Service (CoS)
Translations in IP/MPLS Network”,
[8]. Loa Andersson, Stewart Bryant, "The Proceedings of the 26th Annual IEEE
IETF Multiprotocol Label Switching Conference on Local Computer Networks,
Standard: The MPLS Transport Profile ISSN:0742-1303, page(s):243, 2001.
Case," IEEE Internet Computing, vol. 12,
page(s):69-73, Aug, 2008. [16]. Ji-Feng Chiu, *Zuo-Po Huang, *Chi-
Wen Lo, *Wen-Shyang Hwang and Ce-
[9]. Mahesh K.P, Njulata Yadav S. V, Kuen Shieh, “An Approach of End-to-End
Charhate M.E, “Traffic Analysis of MPLS DiffServ/MPLS QoS Context Transfer in
and Non MPLS Network including MPLS HMIPv6 Net”, Autonomous Decentralized
Signaling Protocols and Traffic distribution Systems, Eighth International Symposium,
in OSPF and MPLS”, 2008. ISBN: 0-7695-2804, page(s):245-254, Mar
2007.
[10]. Srinivas Vegesna Quality of Service
By Published by Cisco Press, 2001 ISBN [17]. Sundeep .B.Singh and Girish.P.Saraph,
1578701163, 9781578701162 “DiffServ over MPLS: Tuning QOS
parameters for Converged Traffic using
[11]. B.E. Nichols, “Differentiated services Linux Traffic Control”, Indian Institute of
in the Internet Carpenter”, Proceedings of Technology Bombay, Powai, Mumbai India.
the IEEE, Volume 90, Issue 9,
page(s):1479–1494, Sept 2002. [18]. Hongyun Man, Linying Xu, Zijian Li,
Lianfang Zhang, “END-TO-END QOS
[12]. Leonardo Balliache, Network QoS IMPLEMENT BY DIFFSERV AND
Using Cisco HOWTO, April 2003 v0.1 MPLS”, Electrical and Computer
COMO1O6-7
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
[21]. John Bartlett and Rebecca Wetzel Sandeep Singh has received his
BCR, “QOS over MPLS–the Complete B. Tech. degree in Electronics and
Story”, P2P for Communications: beyond Communication Engineering from
file sharing, Volume 36, Number 2, Chaudhary Charan Singh University, Meerut
February 2006. in 2008. He is a M.Tech. student at Gautam
Buddha University, Greater Noida. He has
[22]. Victoria Fineberg, “QoS Support in
MPLS Networks MPLS/Frame Relay been worked as a faculty of Dr.K.N.Modi
Alliance”, White Paper May 2003. Institute of Engineering and Technology,
Modinagar, Ghaziabad. He is working as an
[23]. LI Ming-hui and Xiz Jing-bo, Assistant Professor at Sobhasaria
“Research and Simulation on VPN Engineering College, Sikar, Rajasthan. He is
Networking Based on MPLS”, Wireless a life time member of IACSIT, Singapore
Communication, Networking and mobile
and member of CSI. He has published a
Computing ,page(s): 1- 4, Oct 2008.
paper in “national conference on
[24]. Hiroshi Yamada, “End-to-End communication network and security”,
Performance Design Framework of MPLS Alwar, Rajasthan. His research interest
Virtual Private Network Service across includes Ad-hoc wireless networks, Mobile
Autonomous System Boundaries”, IP, Quality of Service, MPLS based
Telecommunications Network strategy and
networks and Traffic Engineering.
planning Symposium , page(s): 1-6, Nov
2006
COMO1O6-9
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
ABSTRACT
Integrated Services Digital Network(ISDN) As a result, ISDN has the ability to deliver a
is a digital communications technology that wide range of desirable applications in
enables a small business or an individual to education, healthcare, business, and
connect directly to both the Internet and multimedia. In light of this substantially
other users. ISDN provides a standard increased functionality, ISDN can be the
interface for voice, fax, video, graphics and basis for a new type of useful medium for
data- all on a single telephone line. ISDN is Internet access that is: reasonably priced,
a set of communications standards for readily available, and also allows for the
simultaneous digital transmission of voice, delivery of information at high speeds, and
video, data, and other network services over provides for many types of new and
the traditional circuits of the public switched previously unavailable applications. The
telephone network. It was first defined in technical design of ISDN is motivated by
1988 in the CCITT red book. ISDN enables the primary aim of modernizing traditional
the telephone network, which was built for telephone networks. The use of digital
traditional analog calls, to carry information switching and transmission technology
digitally at higher speeds and without the within the ISDN facilitates the transmission
errors of the traditional analog system or of digital data besides the digitized voice.
POTs (Plain Old Telephone Services) lines.
KEYWORDS
ISDN, ISDN traffic, voice/data networks, value added services, video conferencing.
INTRODUCTION
Integrated Services Digital Network (ISDN) user/network interfaces, and network and
is comprised of digital telephony and data- internetwork capabilities. ISDN applications
transport services offered by regional include high-speed image applications,
telephone carriers. ISDN involves the additional telephone lines in homes to serve
digitalization of the telephone network, the telecommuting industry, high-speed file
which permits voice, data, text, graphics, transfer, and video conferencing. Voice
music, video, and other source material to be service is also an application for ISDN. This
transmitted over existing telephone. The paper summarizes the underlying
emergence of ISDN represents an effort to technologies and the services associated
standardize the subscriber services, with ISDN. Also, there are two significant
COMO1O7-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
developments drive the commercialization years and a third of these will be fiber optic
of ISDN: digital switching and digital cable. Pulse Code Modulation (PCM) and
transmission. It is estimated that nearly all of Time Division Multiplexing (TDM) are
the access lines and trunk routes in the important related technical developments
United States will be digital within a few enabling digital switching and transmission.
ISDN STANDARDS
The ISDN Basic Rate Interface (BRI) International Telecommunication Union
service offers two B channels and one D Telecommunication Standardization Sector
channel (2B+D). BRI B-channel service (ITU-T) (formerly the Consultative
operates at 64 kbps and is meant to carry Committee for International Telegraph and
user data; BRI D-channel service operates at Telephone [CCITT]) I.430. ISDN Primary
16 kbps and is meant to carry control and Rate Interface (PRI) service offers 23 B
signaling information, although it can channels and one D channel in North
support user data transmission under certain America and Japan, yielding a total bit rate
circumstances. The D channel signaling of 1.544 Mbps (the PRI D channel runs at 64
protocol comprises Layers 1 through 3 of Kbps). ISDN PRI in Europe, Australia, and
the OSI reference model. BRI also provides other parts of the world provides 30 B
for framing control and other overhead, channels plus one 64-Kbps D channel and a
bringing its total bit rate to 192 kbps. The total interface rate of 2.048 Mbps. The PRI
BRI physical-layer specification is physical-layer specification is ITU-T I.431.
ISDN SIGNALING
There are two different types of signaling DSS1 defines what format the data goes in
used in ISDN. For communicating with the on the D-channel, how it is addressed, etc. It
local phone company, ISDN uses the Digital also defines message formats for a variety of
Subscriber Signaling System #1 (DSS1). messages used for establishing, maintaining,
COMO1O7-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
and dropping calls, for instance SETUP handle the signaling needs of ISDN as well
messages, SUSPEND and RESUME as other older signaling systems and
messages, and DISCONNECT messages. (hopefully) will adapt well to future needs.
Once your DSS1 signal makes it the the One important feature of SS7 is providing
phone company, their own signaling system CCS. This makes it harder for malicious
takes over to pass the call information users of the phone network to put one over
within their system, and between other on the phone company. It also improves the
phone companies. Signaling System #7 service, for instance by offering faster
(SS7) is supposed to be used for this. SS7 connection establishment. Older equipment
defines a communications protocol, and still looks for the signaling information in
formats similar to DSS1, however SS7 is the same channel as the voice, in the eighth
designed in a broader, more general way. bit of each piece of voice data.
DSS1 is specific to ISDN, however SS7 will
ISDN SWITCHING
With pure ISDN, switching is the into each device use. The phone company
department of the phone company. then stores this data somewhere, and when
Traditional phone services is Circuit one connect the machine to the network, it
Switched Voice (CSV). The voice data goes sends its SPID to the nearest phone
through several switches before reaching its company switch which identifies what type
final destination. For point-to-point data of connection the device needs and
connections, you need Circuit Switched (therefore) how to route its calls.
Data (CSD) - the exact same thing with data Presumably, the SPIDs have to refer to a
instead of voice. If CSV are used, they are configuration that matches one of the two B-
free to route your call through any type of channels you have. By the way, the SPIDS
switch, even the old analog switches. The are arbitrary numbers that refer to data
digital channel may also be shared with stored by the phone company. The phone
other channels, in the moments when there Company often includes the phone number
is silence on the phone line. And the digital in the SPID for their own convenience. One
parts of a CSV call can go through noisy older type of phone company switch, a
switches that might create an undetected DMS-100, was improperly designed with
error. There are actually a standard set of respect to the standards relating to SPIDs.
combinations defined for setting up BRIs. This switch misguidedly assigns one SPID
These are called National ISDN Interface to each B-channel that is use, rather than to
Groups (NIIGs),so there will be a limited each device. Therefore if the nearest switch
menu of offerings available. One can get is a DMS-100, one will only be able to hook
both B-channels for data, or one for voice up two devices to the CPI, rather than eight.
and the other for data, or one for voice and If one is only going to be hooking up a
the other for either voice or data. In order to single device to the ISDN (i.e. setting it up
facilitate this, North American phone in a point-to-point configuration, one might
companies use an optional part of the ISDN not need a SPID at all, as the phone
standard to identify each TE1 or TA you company can identify the ISDN line as one
use. The phone company assigns a Service particular type, full time. This depends on
Profile Identifier (SPID) to each of these what equipment they have - the old DMS-
devices, and one has to manually enter them 100 switch will still require to have a SPID.
COMO1O7-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
ISDN FRAME
The usual reference to an ISDN implies an a communications node with circuit traffic.
integration of services on the sense of the The potential for combining the
transport of digitalized voice and data, the telecommunications and the computer
former with telephonic signaling and the technologies becomes apparent when the
latter either gaining access via telephonic delivered signal is not just a replica of the
signaling and then employing data signaling input signal, but has been processed in a
or via data signaling only. Circuit and useful manner. It can be debated on
message, synchronous and asynchronous, operational, economical, technical and
packet - all traffic types are to be political grounds whether such processing
accommodated. Packet switching is should be included within the public and
considered by some as a value-added private network itself, treated and offered as
service, but by others as a transport a network service, or whether these
mechanism, which can be combined within functions should be external to the network,
provided at its periphery.
COMO1O7-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
ISDN VIDEOCONFERENCING
Today, a modern organization is connected meetings. When at their desks, employees
via an IP network resembling a “hub” have both Intranet services and access to the
between dedicated connections in multiple public Internet. Videoconferencing is being
remote offices. Employees use the IP explored as a technology that can improve
network for text-based communications, group and individual communications and
meet regularly via telephone and frequently thereby affect both these strategic
travel to remote offices for face-to-face objectives.
CONNECTIVITY OPTIONS
In the past, network options for the local exchange carrier. As a result of the
videoconferencing were simpler. Heavy improved ability to guarantee quality of
users placed endpoints on a common services in packet networks, the ubiquity of
“private network” of switched circuits switched local area Ethernet networks, and
leased from telecommunications carrier. the release of numerous products for
Most companies preferred to connect business quality H.323 videoconferencing,
videoconferencing endpoints to a public IP-based networks are now firmly on the list
network via Integrated Switched Digital of videoconferencing connectivity options.
Network (ISDN) connections obtained from
Data Storage:-
additional cost and work, these systems are
There are two types of videoconferencing
not very well suited for secure
systems, or codecs, commercially available
videoconferencing.
today; PC-based and appliance.
APPLIANCE VIDEOCONFERENCING
The PC-BASED VIDEOCONFERENCING
SYSTEMS are not based on a PC platform.
SYSTEMS that run PC operating systems
These devices have been custom designed
and utilize, at least in part, standard COTS
and manufactured to provide only specific
PC hardware. Since these systems include
functionality and typically do not utilize
internal hard drives and plug-and-play
standard PC- based hardware or software. In
connectivity for external storage devices,
addition, the storage capabilities of these
one must take additional steps to secure
systems are usually limited to storing
these systems. Specifically, the system must
address book information, usage data, and
be equipped with a removable hard-drive
configuration settings. Since these devices
setup. In addition, two different hard drives
do not provide data storage capabilities, they
must be used; one for secure calls and one
are a good choice for secure
for non-secure calls.Because securing a PC-
videoconferencing.
based videoconferencing system requires
COMO1O7-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
BENEFITS OF VIDEOCONFERENCING
ISDN BACKUP
COMO1O7-6
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
APPLICATIONS OF ISDN
Studio quality audio transmission for Preparation of printed materials.
broadcast. Electronic manuals online.
Image archives(Real estate, medical Stock quote for brokers.
images, photographic image banks etc.). Credit card authorization.
CONCLUSION
As ISDN is deployed, few people are would be the Pipeline 25, from Ascend,
currently replacing their home phone system which provides ISDN to Ethernet
with an ISDN phone network. The trend for connections, using IP. It has an NT1 built in,
now seems to be providing an entire ISDN and provides two phone jacks for standard
network in a single box, with the NT1, TA, POTS telephones.
and TE1 equipment all built in. An example
REFERENCES
COMO1O7-7
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Abstract: Wired network is an interconnection secure and long haul communications. The wired
of remote nodes through physical networks are most suitable for environments with
infrastructure of wires and cables. It provides fixed entities like home and office networks and
advantages of fast, reliable, secure and long high data rates environments where dedicated
haul communication. The modeling and links are necessary as a link between keyboard
analysis of a network provides understanding and central processing unit (CPU). The primary
of behavior of various network parameters parts of a wired network are network cables,
which can be further used for computation of network adapters, hubs, switches, routers and
desired results and efficiency. In this paper process governing software’s like transmission
analysis of wired networks has been done to control protocol/internet protocol (TCP/IP), token
evaluate the performance of a network under ring, fiber distributed data interface (FDDI). The
varying bandwidth of transmission links and processing capabilities and methodologies of the
number of workstations (nodes) attached to the computer hardware and networking components
network. may be slightly varied by using different network
processing software’s. Institution of electrical and
Keywords: OPNET, Wired Networks, electronics engineering (IEEE) model IEEE 802.3
Transmission links Ethernet working on transmission control
protocol/internet protocol (TCP/IP) model is most
I. INTRODUCTION famous of all the wired local area network models.
Data communication and networking is the fastest The most common Ethernet links used are
growing technology in the modern era. Advanced 10BaseT, 100BaseT and 1000BaseX providing a
research projects agency network (ARPANET) data transfer rate of 10 Mbps, 100 Mbps, and 1
added fuel to the growth of wired network Gbps respectively. Fiber optic variants of Ethernet
industry and since then wired networks are offer high performance, electrical isolation and
emerging as social networks which are linking network span up to tens of kilometers. The
people, organizations and knowledge worldwide. backbone of internetwork of interconnected
This radical change in the wired networking networks (INTERNET) relies on the huge
domain over the last decades is the result of worldwide established wired network known as
convergence of internet coupled with engineering public switch telephone network (PSTN). Some of
advances in the field of information and the limitations of wired networks include
communication technologies. Wired computer immobility, uneasy upgradation, difficult fault
networks have bridged the gap between countries diagnosis, elaborate infrastructure, intricate
and continents and brought up an era of installation and its non availability at isolated
globalization. places.
COMO1O8-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
components should facilitate performance, mainly used in long haul communications and
reliability and security of the network. The high speed local area networks (10BaseF).
computer network component includes
transmission media, networking devices Networking Devices
(hardware) and network governing software’s. Networking devices or hardware devices or
intermediate units (IU) are equipments that
Transmission Media mediate data in a computer network. Networking
Transmission media is a physical link between the devices play a vital role in describing the behavior
network nodes that connects the nodes to form a and characteristics of network parameters like
network. Physical links may vary from a common delay, transmission rate, throughput and
twisted pair cable to an expensive fiber optic cable bandwidth. Common networking devices are
according to the scale and performance criterion Network Interface Cards (NICs), modulator
of the network. Twisted pair cables consist of two demodulator (MODEM), Bridges, Hubs, Switches
conductors (normally copper) each with its own and Routers.
plastic insulation twisted together. The twisting of Network interface cards (NIC) provides
cables together eliminates unwanted signals like physical access to a network media (link)
interference and crosstalk. The most common through the use of medium access control
twisted pair cable used in communication is (MAC) addresses. Network interface cards
referred to as unshielded twisted pair (UTP). The (NICs) operates at a speed ratings of 11
electronic industries association has developed Mbps, 54 Mbps and 100Mbps.
standards to classify unshielded twisted pair cable Repeater is an electronic device that
into seven categories, defining the cable quality receives a signal cleans it of unnecessary
with 1 as the lowest and 7 as the highest. Twisted noise, regenerates it and transmits it at a
pair cables are used in telephone lines to provide higher power level, so that the signal can
voice and data channels. Digital subscriber line cover longer distances without
(DSL), Ethernet (10BaseT, 100BaseT) employs degradation. In most twisted pair Ethernet
twisted pair cables as their physical media. configurations repeaters are required for
Coaxial cables as transmission media are used to cable that runs longer than 100 meters.
carry signals of higher frequency ranges. These Network hub or concentrator duplicates
cables carry a core conductor of copper wire the data packets received through one port
enclosed in an insulating sheath which in turn is and makes it available to all the ports
encased in an outer conductor of metal foil. attached to it allowing data sharing
Coaxial cables are employed in television and between all devices connected to the hub.
telephone networks. The network switch is a device that
forwards datagram between ports based on
the MAC addresses in the data packets.
Switch is distinct from a hub in that it only
forwards the frames to the ports involved
in the communication rather than to all
ports connected.
A router is used to connect subnets to the
main network or two distinct networks
together, provided they are based on the
same network model. As router only
Figure 1 Twisted pair cable and coaxial cable
understands the addressing modes of
similar networks (networks that are
A fiber optic cable is made of glass or plastic and
governed by same protocol suite).
transmits signals in the form of light. It provides
enormous amount of bandwidth and transmits data Gateway is a device that is used to transfer
over large distances. Optical fiber cables are data between dissimilar networks
COMO1O8-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
(networks that are governed by different network operations such as World Wide Web
protocol suite). (WWW), electronic-mail, instant messaging,
Bridge is used in packet-switched voice over internet protocol (VoIP) and network
computer networks and unlike routers; based audio and video services thus making
bridge makes no assumptions about where computer networks an important part of
in a network a particular address is infrastructures for home and offices. Therefore it
located. It depends on flooding and is often desirable to improve the efficiency and
examination of source addresses in designing of computer networks. Analytical
received packet headers to locate unknown methods using queuing network theory and
devices. Once a device has been located, computer simulation methods are extensively used
its location is recorded in a table where the for network analysis and evaluation. Former is a
MAC address is stored. Their application conventional approach used for smaller networks
is limited to local area networks because of and later is best suited for the analysis of complex
its lengthy algorithms. and large networks of nowadays.
The modeling and analysis of a network provides
the understanding of the behavior of the network
parameters which can be used for analysis of
desired results and higher efficiency. Moreover
the study helps in increasing the application
domain of the technology and sometimes may also
result in discovery of newer technologies.
V. INTRODUCTION TO PROBLEM
A comparative study of network parameters
delay-global (sec), delay-node (sec), load
(bits/sec) and traffic received (bits/sec) is made by
varying the physical characteristics and logical
conditions of computer networks. Physical
characteristics involved number of nodes and type
Figure 2. A wired computer network of cabling. The size of data packet is chosen from
logical domain. The delay may be measured as a
IV. COMMUNICATION SOFTWARES global parameter or as a nodal parameter. In
A network uses number of protocols to establish a global sense, the delay may be defined as an
link between nodes for data and resource sharing. average time gap experienced by all the nodes in a
The most common protocol that governs the network between transmission and reception of
communication in almost every network is data packets or between request and response. The
transmission control protocol/ internet protocol delay in nodal sense is defined as the average time
(TCP/IP) known as the internet protocol suite. taken by the data packets to reach a particular
Some other important protocols working under or node. It is measured in seconds. Load is the
over the internet protocol suite includes hypertext amount of data packets transmitted by a particular
transfer protocol (HTTP), file transfer protocol node, thereby adding load on network. It is a
(FTP), medium access protocol (MAC),address nodal characteristic and is measured in bits per
resolution protocol (ARP), secure file transfer second (bps). Traffic received is the amount of
protocol (SSH), internet control message protocol throughput measured for the particular node. It is
(ICMP) and multi protocols label switching measured in bits per second or packets per second.
(MPLS). The arrival of internet work of The comparison of delay, traffic received and
interconnected networks (INTERNET) into daily load is computed by varying
life has greatly influenced the relevance and Type of link which may be categorized
importance of networks. Nowadays more and depending upon its speed of carrying data.
more applications rely on efficient and reliable
COMO1O8-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO1O8-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
On changing the cable from 10BaseT, The results obtained by simulation of the network
100BaseT to 1000BaseX, no variation in verifies that the Ethernet cables are categorized on
load pattern is observed across the node the basis of data rates.10BaseT, 100BaseT and
(node_0) as recorded by the simulation 1000BaseX categories of cables that provide
results. different data rates. 10BaseT is the slowest of two
and offers maximum delay, whereas 1000BaseX
is fastest and has negligible delay.
COMO1O8-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO1O8-6
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
network with 10 nodes attains the least [5] C.S. Raghavendra, M. Gerla, A. Avizienis,
linear value. “Reliable loop topologies for large local
computer networks”, IEEE transactions on
computers, January 1985, vol. c-34, issue
Table 2 Result comparison: Network 2
1, pp 46-55
NETWORK ATTRIBUTES NETWORK PARAMETERS [6] Shenker, David Clark, Deborash Estin,
Type of
No.
of
Packet
Delay-
Global
Load
Traffic
Received
Shai Herzog, “Pricing in computer
cable
node
Size
(Sec)
(bps)
(bps) networks: reshaping the research agenda”,
10BaseT 10 1500 19 520 560 Telecommunication policy, April 1996,
millisecond approx. approx.
s vol.2, issue 3, pp 183-201
100Base 20 1500 16 620 590 [7] Laurent Paquereau, Bjarne E. Helvik, “A
T millisecond approx. approx.
s module based wireless node for NS2”,
100Base 30 1500 15 625 600 p Proceedings from the workshop NS2: The
X millisecond approx. r
s o IP network simulator, 2006, article 4
x [8] Qichen, Felix Schnidt- Eisenlohr, Daniel
Jiang, Marc-Torrent-Moreno, Luca
Delgrossi, Hannes Hartenstein, “Overhaul
VIII. CONCLUSION of IEEE 802.11 modeling and simulation
The Ethernet cables are categorized on the basis in NS2”, Proceedings of the 10th ACM
of data rates.10BaseT, 100BaseT and 1000BaseX symposium on modeling, analysis and
categories of cables that provide different data simulation of wireless and mobile systems,
rates. 10BaseT is the slowest of two and offers 2007, pp 159-168
maximum delay, whereas 1000BaseX is fastest [9] Min-Seak Kang, Jaeshin Jang,
and has negligible delay. The delay and traffic “Performance evaluation of IEEE 802.16d
received increases with increase in the network ARQ algorithm with NS2 simulator”,
load. Communications magazine, August 2006,
pp 1-5
REFERENCES [10] D. Mahrenholz, S.Ivanov, “Real time
[1] S. Pierre, G. Legault, “A genetic algorithm network emulation with NS2”, Eight IEEE
for designing distributed computer International symposiums on distributes
network topologies”, IEEE transaction on simulation and real time applications,
system, man and cybernetics, April 1998, October 2004, pp 29-36
vol. 28, issue 2, pp 249-258 [11] S. Kurkowski, T. Camp, N. Mushell, M.
[2] F. Kamoun, L. Kleinrock, “Analysis of Colagrasso, “Visualization and analysis
shared finite storage in a computer tool for NS2 wireless simulations:
network node environment under general iNSpect”, 13 IEEE International
traffic conditions”, IEEE transaction on symposium on modeling, analysis and
communications, July 1980, vol. 28, issue simulation of computer and
7, pp 992-1003 telecommunications systems, September
[3] Yu-Chee Tseng, Chih-shu Hsu, Ten- 2005, pp 503-506
Yueng Hsieh, “Power saving protocols for [12] Li Huijie, Lin Xiaokang, “A network
IEEE 802.11 based multi hop adhoc based 3 tier network simulation
networks”, computer networks magazine, architecture”, 2005 IEEE International
October 2003, vol.43, issue 3, pp 317-337 symposium on communication and
[4] B.C. Neuman, T. T’so, “Kerberos: an information technology, October 2005, pp
authentication service for computer 793-796
networks “, IEEE communication [13] Yijinzhang, Pingping Xu, Guangguo Bi,
magazine, September 1994, vol.32, issue Sheng Bao, “Analysis of energy efficient
9, pp 33-38 power saving in IEEE 802.15.4”, IEEE
COMO1O8-7
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO1O8-8
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
MARCH
2011
COMO1O9-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO1O9-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
as,
1) Continuous-time PAPR
max0≤n≤N L−1 |s(n)| 2
In general, the PAPR of OFDM signals is de- P AP R[s] = (4)
fined as the ratio between the maximum in- E|s(n)|2
stantaneous power and its average power where denotes the expectation operator.
2) Discrete-time PAPR
Fig. 1 shows the distribution of the PAPR A. Definition and Remark : Let f (.), be
of the OFDM signals with N=256 and a function and let s(t) = |s(t)| exp jϕ(t) be
L=1,2,4,16. As shown, the largest PAPR in- a single or multicarrier signal, where ϕ(t)
crease happens from to L=1 to L=2. However, is the phase of s(t). The function f (.) is
the PAPR does not increase significantly after a function of s(t) PAPR reduction if the
L=4. It has shown that L ≥ 4 is sufficient PAPR of s(t) is strictly inferior to the PAPR
to get accurate PAPR results . The PAPR of s˜(t) = f [|s(t)|] exp jϕ(t), i.e., it exists
computed from the L-times oversampled time ∆P AP R > 0 such as
domain OFDM signal samples can be defined
COMO1O9-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
P AP R[s˜] + ∆P AP R = P AP R[s](indB),
C. PAPR reduction techniques based on
where ∆P AP R is the gain in PAPR re- Nonlinear Functions in adding signal context
duction.Indeed, a signal s(t) and its
scaled version have the same PAPR, i.e., The adding signal techniques consist of
P AP R[s] = P AP R[αs], where α,is a scalar. reducing the envelope of OFDM signal by
adding a peak-reducing signal[5] just before
B. Proposition the HPA as shown in Fig.1
The uncorrelated component resulting in non-
linear process of the input signal amplitude is
the component for PAPR reduction. Proof:
Let f (.) be a function of s(t) PAPR reduction,
s(t) can be a single or multicarrier signal.
P AP R[s˜] + ∆P AP R = P AP R[s](indB),
Figure 2: Adding signal scheme for PAPR re-
(ii) As f(.) is a nonlinear function (refer to duction,where c(t) is the peak reducing signal
the remark), using the Bussgang decomposi-
tion[4], the signal s̃(t) = f [|s(t)|] exp jϕ(t)can
be written as The PAPR reduced signal is therefore
expressed by
s̃(t) = αs(t) + d(t),
s˜(t) = s(t) + c(t)
where α = Rs̃s (0)/Rss (0). The compo-
nent d(t) is uncorrelated with the input signal from Eq and Eq , the peak reducing sig-
s(t), i.e.,Rsd (τ ) = E[d(t + τ )s̃(t)] = 0. nal c(t) is expressed as
(iii) As, s(t) and its scaled version αs(t) have c(t) = (α − 1)s(t) + d(t)
the same PAPR, i.e., P AP R[s] = P AP R[αs],
and as from above Eq we can see that ,the peak
reducing signal depends on the uncorrelated
P AP R[αs+d]+∆P AP R = P AP R[s̃]+∆P AP R component resulting in the nonlinear process
(5) of the input signal amplitude.
= P AP R[s] (6)
= P AP R[αs]; (7) D. Properties of the peak-reducing sig-
nal
(iv) Therefore,
COMO1O9-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
the correlation function for c(t) is Here we use the unused subcarriers em-
bedded int he spectrum of multicarrier
Rcc (τ ) = |α − 1|2 Rss (τ ) + Rdd (τ ) (9) standards to carry the peak-reducing signal
in order not to degrade the BER of system
where ,Rdd (τ ) is the correlation function for and notto generate OOB distortions.Without
the uncorrelated component d(t) and is given unused subcarriers of multicarrier standards
by ,PAPRcan be reeuced without BER degra-
∞ dation but with OOB emission or without
X
Rdd (τ ) = bn [Rss (τ )/Ps ]( 2n + 1) (10) OOB emission but with BER degradatoin. A
n=1 FFT/IFFT pair-based digital filtering to filter
the peak reducing signal in order to carry it
and only on the unused subcarriers of multicarrier
Z standards. Let I and O be be the set of the
1 1 r2
bn = k rf (r)p(r)LnI (
)dr k in band and out band indices respectively
Ps n + 1 Dr ps
(11) and {i0 , ...., iNr −1 } be the locations of the
whereDr = r : 0 ≤ r ≤ ∞ is the domain of unused subcarriers and R c the complement
integration,p(r) is the probability density of of R in I. The forward FFT transforms cn
the input signal amplitude andLIn (x) is the back to the frequency domain.The discrete
languerre function expressed by frequency components of cn on the unused
subcarriers R are passed unchanged while the
x−k exp x data subcarriers Rc and the OOB components
LIn (x) = (d/dx)n (xn+k exp −x) O are setted to zero, i.e,
n!
(12) (
The peak reducing signal power spectrum Ck , k<
C˜k =
density is expressed by 0, k (<c ∪ O)
P∞
n=1 Ps2n+1 [Sss (ϑ) ⊗1
bn
Scc (ϑ) = |α − 1|2 Sss (ϑ) +
.... ⊗2n+1 Sss (ϑ)]
COMO1O9-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
signal cn at the output Tof the filter-based amplitude of the signal. In [11],it is shown
FFT/IFFT. Because of R Rc = Φ, the BER that any form of clipping (in particular, the
of the system is not degraded. Because of classical clipping) can be formulated as an
C̃k = 0 for k O, there is not OOB emission. adding signal technique as illustrated in Fig.1.
It is obvious that the filtering eliminates a
part of the peakreducing signal which causes B. The IEEE 802.11a/g standards based
a peaks regrowth. In order to reduce as much WLAN system
as possible the PAPR, the process of adding
signal computation followed by filtering must In WLAN IEEE 802.11a/g standard IFFT
be repeated several times. The new PAPR size (N) is 64. Out of these 64 subcarriers, 48
reduction scheme including the filtering and subcarriers are used for data, while 4 subcarri-
the iterations is illustrated in Fig.2 in discreteers are used for pilots. The rest 12 subcarriers
time-domain. are unused (null) subcarriers located at the
positions R = {0, 27, ......, 37} of the IFFT
VI. ILLUSTRATION BY THE CLASSI- input. Only these unused subcarriers shall be
CAL CLIPPING FUNCTION IN A WLAN utilized for WLAN PAPR reduction.
SYSTEM CONTEXT Fig.3 shows the IEEE 802.11a/g Standard
specifications: Subcarriers (data, pilots, un-
There are many functions which can be used) positions [9].
used for PAPR reduction such as clipping
,companding techniques and within compand- VI. CONCLUSION
ing there exist exponential companding , µ
-law compander,A-law compander, exponen- OFDM is a very attractive technique for
tial companding and many more.In this paper wireless communication due to its spectrum
, we will discuss the classical clipping function efficiency and channel robustness.One of the
which is a function for PAPR reduction.This serious drawback of OFDM system is that
function is used to reduce the PAPR of the the composite transmit signal can exhibit a
local-area-networks (WLAN) system based on very high PAPR when the input sequences
IEEE 802.11a/g standards under the condition are highly correlated.In this paper using
of no BER degradation and no OOB power Bussagang theorem we have pointed out that
emission. input signal when undergoes nonlinear process
results in an uncorrelated component which
A. Clipping function is used as a component for PAPR reduction
Using the conventional clipping technique [8] and unused subcarriers in the spectrum of
to reduce OFDM PAPR, the output signal uncorrelated component is used to carry
s˜(t), in terms of the input signal s(t) is given the component for PAPR reduction.Here
as follows: s̃(t) = f [|s(t)|] exp jϕ(t), where a conventional clipping technique has been
ϕ(t) is the s(t) phase and f(.) is the clipping formulated as an adding signal technique
function which is expressed as for PAPR reduction and has been used to
( illustrate these assertions in a WLAN system
r, r≤A based on IEEE 802.11a/g standard.
f (r) =
A, r≥A
REFERENCES
where A is the clipping threshold and r is the [1] Y.Wu and W. Y. Zou, ”Orthogonal frequency
COMO1O9-6
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO1O9-7
Analysis on Placement of Wavelength Converters in WDM p-Cycle Network
The p-cycle concept combines the merit of the two 4.1. Partial wavelength conversion
basic protection strategies so far known: ring
protection and mesh restoration. On one hand, it WDM networks add the aspect of wavelength
allows short recovery times of about 50-150 ms known assignment to the p-cycle protection concept. Working
from protection ring structures. On the other hand it and p-cycle links may have different transmission
has been repeatedly validated that under appropriate wavelengths. It is not sufficient to provide only
design methods, it offers the capacity-efficiency which enough p-cycle protection paths for the working links
is essentially as high as that of a span-restorable mesh but its also have to consider the wavelengths of the
network. working paths failed in any given scenario and the
wavelength(s) on which a p-cycle is established. If the
3. Wavelength converters in p-cycle networks protection path for a working link is allocated at a
different wavelength, the wavelength must be
When wavelength conversion is not available, a converted to access the p-cycle in case of a failure.
lightpath must use the same wavelength on all the Otherwise, all working link and protection path
links traversed in a WDM optical network. This arrangements must be coordinated to have the same
requirement is known as the wavelength continuity wavelengths.
constraint. On the other hand, if wavelength routers In nodes with partial wavelength conversion,
are capable of wavelength conversion, an optical only a limited number of incoming lightpaths can
signal may be converted from one wavelength to change to a different wavelength on the outgoing link.
another wavelength. In some previous work on -cycle Fig 4.1(a) and 4.1(b) depict an optical cross-connect
based protection, a path following the wavelength node with a shared pool of C wavelength converters. If
continuity constraint is called a wavelength path (WP), there is no wavelength conversion required, an
while a virtual wavelength path (VWP) is defined to be incoming lightpath will be directed to the appropriate
a path that uses wavelength conversion at each node output port of an outgoing fiber or to the local access.
Otherwise the lightpath can pass through an available Advantages of partial wavelength conversion:
converter in the converter pool. As in a WP network 1. No. of wavelength converters is reduced.
there are no converters, and in a highly equipped 2. Blocking probability of calls is reduced i.e. success
VWP network each node has at least m*n converters, probability is increased.
where m is no. of input and output fibers in the optical
switch and n is the total no. of wavelengths in 4.2. Sparse wavelength conversion
wavelength mux and demux.
With this type of node architecture in mind, it
is considered to have two basic alternatives to provide To achieve the full wavelength conversion capability,
p-cycles in WDM networks with partial wavelength i.e. the input wavelength can be converted into any
conversion. One employs WP p-cycles with converters output wavelength; a simple method is to use a
used for WP working paths to access them, and the converter per wavelength per port in a dedicated
other is based on VWP p-cycles. The first idea is that manner. However, from commercial point of view this
in order to provide for protection without requiring a method is not cost effective as the cost of wavelength
set of p-cycles dedicated to every wavelength, but converters is quite high and also the quality of signal
while using as few converters as possible overall, it is also degraded. So it is recommended to use the
may be efficient to associate converters only with the minimum no of converters. To reduce the usage of
p-cycle access points-leaving the working paths to be wavelength converters, there are two approaches. In
implemented in a pure WP manner incurring no one approach, sharing of wavelength converters is
converter costs. Although in practice the converters done through a switch as discussed in the aforesaid
may be available for working paths as well, it is approach. Or in another approach, wavelength
assumed for research comparisons on the basic idea converters are allocated to only few of the nodes in the
that the converters are used for protection paths only. network, i.e. some of the nodes possess the wavelength
One reason for this is also that protection paths conversion capability while others do not. This refers
typically need more flexibility in the wavelength to sparse wavelength conversion. Also the sparse
selection to be able to protect many working wavelength conversion fulfils the requirement of
wavelengths. A built-in advantage is also that because wavelength conflict rule. According to wavelength
a failed working link is replaced by a longer protection conflict rule, the number of wavelengths required in a
path, the converters (which are typically o-e-o WDM network is at least equal to the maximal
regenerative circuits) can ensure that the signal quality number of channels over a fiber (called maximal link
of affected paths is not degraded by the protection re- load) in the network. By placing wavelength
routing. In this basic approach, after detecting a span converters at some nodes in the network (i.e. sparse
failure the failure-adjacent nodes switch the working wavelength conversion), the number of wavelengths
link to a predetermined protection path on a p-cycle needed can be made equal to the maximal link load.
and, if necessary, convert the wavelength of the The converter allocation principle is explained with
working link to that of the p-cycle at the access point. the help of figure 4,4(a) and 4(b). In figure 4 there are
two working paths A-B-C and B-A-E and the p-cycle
configuration has been completed in the network. In
the figure, the shaded nodes indicate the places where
converters are needed and the numbers associated with
each edge represent the wavelengths assigned to each
link passed through the working paths and p-cycles.
6. References
Abstract: - The traditional wired transmission medium poses Reliability: The wireless networks are subject to interference
constraints like mobility and extensive cabling. As the networks are and can thus pose a problem in the administration of Wireless
being upgraded from scratch all over the world, network planning Infrastructure.
is becoming all the more important. Computing the viability and Security: Technically wired LANs are more secure than
performance of computer networks in real can be very expensive WLANs. Since wireless signals are transmitted through the air,
and painstaking task. To ease and comfort the process of
they can be captured by devices outside the network. However,
estimating and predicting a network design, simulation and
modeling techniques are widely used and put into practice. Among the majority of wireless LANs today protects their data with the
the variety of simulation tools available OPNET was used for the Wired Equivalent Privacy (WEP) encryption standard or Wi-Fi
purpose of modeling and simulation. In wireless networks, the Protected Access (WAP) which makes wireless communications
metrics like delay, retransmission attempts and throughput have almost as safe as wired ones in homes.
been estimated with varying number of nodes, physical Bit Error Rate (BER): The wireless network’s media is error
characteristic, buffer size, data rate, RTS threshold and prone hence its BER is higher than the wired LANs.
Fragmentation thresholds. Carrier Sensing: Carrier sensing is difficult in wireless
Keywords: OPNET, Buffer size, RTS, FTS, Data rate. networks because a station is incapable of listening to its own
transmissions.
I. INTRODUCTION Hidden Terminal Problem: The hidden terminals decrease the
Due to the various drawbacks of wired LANs like performance of the wireless LANs.
extensive cabling and immobility etc., the wireless technology
gained momentum. Wireless communication technologies II. PERFORMANCE ANALYSIS
employ infrared, spread spectrum and microwave radio Here, the simulation scenario of the wireless LAN has
transmission techniques with varying data rates. The demand of been described in figure 2.1. A snapshot of the wireless LAN
wireless LAN has increased over a span of time because of its models is shown here:
comparative simplicity, flexibility, high rate access and low
cost. The wireless network infrastructure is useful to provide
accessibility in rough terrains and even rural areas where
establishing wired infrastructure is difficult.
COMO2O1-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
packet the delay is recorded when the packet is sent to the objects variation may occur in terms of number of stations, data
physical layer for the first time. The smaller Slot Time and SIFS rate and type of network load too among certain other
values can decrease the average media access delay, and hence parameters. The phenomenon of retransmissions as shown in
improve the performance of the wireless network. The increase Figure 3 is due to heavy congestion in the network or failures,
in delay leads to retransmissions as shown in Figure 3. Total which lead to an excess in maximum TCP retransmission
number of retransmission attempts by all WLAN MACs in the timeouts. The effect of change in physical characteristics on
network occurs until either packet is successfully transmitted or throughput i.e. on the bit rate sent to the higher layer is shown in
it is discarded as a result of reaching the short or long retry limit. Figure 4.
The retransmission attempt counts recorded under this statistic
Case 2: Performance Analysis for varying number of
also include retry count increments due to internal collisions. A
workstations
higher level of offered load brings a large number of collisions.
This section presents simulation analysis under the
influence of offered load, for ERP- 802.11g, in form of
increased number of stations. Throughput and Delay as a
function of offered load for different number of stations have
been realized. The network has been configured to 11 Mbps
medium capacity. All other parameters are according to the
standard specifications.
Table 3 Performance Analysis scenarios for varying number of workstations
Parameters Case 2.1 Case 2.2 Case 2.3
Number of
3 6 12
Workstations
COMO2O1-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Figure 5 Media Access Delay for varying number of workstations Figure 7 Throughput for varying number of workstations (node level)
The throughput/delay can vary due to collisions of data The analysis showed that the throughput (bits/sec) at
packets. When a collision occurs, the received packets are the client node reduces with the increase in the number of
discarded without recovering any data. New transmissions of the workstations in the network. The statistic Throughput (bits/sec)
same packets must follow, possibly inducing secondary –a Node Statistics: means the total data traffic in bits/sec
collisions, so that more slots are used than transmitted packets. successfully received and forwarded to the higher layer by the
Therefore, throughput decreases and delay can become WLAN MAC. This statistic does not include the data frames
excessive. High delay variability is also due to retransmissions that are 1) unicast frames addressed to another MAC, 2)
at the link level. One of the reasons the retransmission occurs is, duplicates of previously received frames, and 3) incomplete,
when ACK are not available in time, due to heavy delay or meaning that not all the fragments of the frame were received
congestion in the network. Link-layer retransmissions induce within a certain time, so that the received fragments had to be
delays which do not conform to the assumptions on which the discarded without fully reassembling the higher layer packet.
transport protocol is based. This causes undesired TCP control The network load applied increases with increase in number of
actions which reduce throughput. Through simulations, it has stations. The retransmission attempts and hence the delay in the
been found that the effective throughput of wireless networks increased number of nodes is also more. The reduction of the
decreases as the retransmission limit increases as shown in maximum of realized throughput with the increased number of
Figure 5. contending stations is very characteristic. A higher level of
offered load brings a large number of collisions. The lack of
RTS/CTS mechanism causes a high level of losses that arise
from a large number of relative large data frames. The graph in
Figure 7 shows that increase in delay occurs due to the offered
load, lower is the throughput for the particular scenario. Another
Simulation scenario focuses on such methods like increasing the
buffer size or reducing the network load by providing restricted
access to the users.
Case 3 Performance Analysis for varying Buffer Size
Buffer size (bits) specifies the maximum size of the
higher layer data buffer in bits. Once the buffer limit is reached,
the data packets arrived from higher layer will be discarded until
some packets are removed from the buffer, so that the buffer has
Figure 6 Retransmission Attempts for varying number of workstations some free space to store these new packets.
An analysis of the presented results in figure 5 (shows Table 4 Performance analysis scenarios for varying Buffer size
node statistics for throughput) allows drawing the conclusions Parameters Case 3.1 Case 3.2
regarding the impact of number of stations on throughput. Number of Workstations 12 12
RTS Threshold(bytes) None None
Fragmentation Threshold(bytes) None None
Data Rate (bps) 11 Mbps 11 Mbps
Physical Characteristics Direct Sequence Direct Sequence
COMO2O1-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO2O1-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Table 5 Performance Analysis scenarios for varying Data Rates the number of bits received increases. Also the media access
Parameters Case 4.1 Case 4.2 Case 4.3 Case 4.4 delay reduces with the increases in the data rate. It occurs with
Number of increase in data rate the data will stay in the buffer for lesser
12 12 12 12
Workstations time and hence the number of retransmissions will decrease.
RTS This indicates that that more number of packets can be delivered
None None None None
Threshold(bytes)
safely with increase in data rate.
Fragmentation
None None None None
Threshold(bytes) Case 5 Performance Analysis for varying RTS Threshold
Data Rate (bps) 1 Mbps 2 Mbps 5.5Mbps 11Mbps
Physical ERP ERP ERP
RTS technique is used to reduce the probability of two
ERP
Characteristics (802.11g) (802.11g) (802.11g) (802.11g) stations colliding with each other. The collisions occur because
Buffer Size (bits) 256000 256000 256000 256000 they cannot hear each other due to the absence of medium
The Figures 11 shows the average media access delay (wired link) between them. In this mechanism, a station wanting
and figure 12 shows the average throughput for varying data to transmit a packet first transmits a short packet called RTS,
rates. The figure 11 shows that the delay is relatively more for which includes the source, destination and duration between the
data rates ( 1Mbps), whereas it is less for scenarios with high packet sent and the acknowledgement received. The destination
data rates (11 Mbps). station responds (if the medium is free) with a response control
packet called CTS (clear to send), which includes the same
duration information. The mechanism reduces the probability of
a collision. The reduction in collision reduces the overhead of
collisions.
Table 6 Performance Analysis scenarios for varying RTS threshold
Parameters Case 5.1 Case 5.2 Case 5.3
Number of
12 12 12
Workstations
RTS
None 256 1024
Threshold(bytes)
Data Rate (bps) 1 Mbps 1 Mbps 1 Mbps
Physical Frequency Frequency Frequency
Characteristics Hopping Hopping Hopping
Buffer Size (bits) 256000 256000 256000
Figure 11 Media Access Delay for varying Data Rate The RTS or the CTS overhead are short in size as
The high data rate provides more bandwidth, so the compared to the data packets to be transmitted. Moreover the
data packet utilizes this bandwidth to be transmitted at high RTS and CTS frame size is controlled per station by a parameter
speed. Thus the delay increases for low data rates called RTS Threshold.
whereas it is less for high data rates (Figure 12).
Figure 12 Throughput (bits/sec) for varying Data Rate (Global) Figure 13 Media access delay for varying RTS
From the above simulation scenarios it has been The increase in delay for high value of RTS threshold
observed that as we increase the data rate form 1 Mbps to 11 shows the increase in delay in Figure 13. Though RTS helps to
Mbps, there is increase in throughput. This is also predictable reduce the collisions occuring due to hidden nodes but the delay
from the theoretical viewpoint that as we increase the data rate, increases due to the additional overhead introduced by thehigh
COMO2O1-6
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
value of RTS threshold . So, RTS threshold value must be bytes long). The mechanism of fragmentation divides the large
optimally chosen. sized packets in to smaller packets called fragments. The need
of fragmentation i.e. smaller packets in a wireless LAN
environment occurs because of the high Bit Error Rate of a radio
link and the probability of a packet getting corrupted (it
increases with the packet size but if the data packet size is
smaller, the overhead to transmit the data packet is also less).
Table 7 Performance analysis scenarios for varying Fragmentation threshold
Parameters Case 6.1 Case 6.2 Case 6.3
Number of
12 12 12
Workstations
RTS
None None None
Threshold(bytes)
Fragmentation
None 256 1024
Threshold(bytes)
Data Rate (bps) 11Mbps 11 Mbps 11Mbps
Figure 14 Retransmission Attempts for varying RTS Physical Direct Direct Direct
Characteristics Sequence Sequence Sequence
The number of retransmission occur with an increase in Buffer Size
256000 256000 256000
number of collisions (Figure 14). The RTS standard reduces the (bits)
collisions, so, the retransmissions occuring due the RTS So, the fragmentation/reassembly mechanism at the MAC Layer
threshold value 1024 bytes decreases as compared to the has been used to deal with large sized packets. Table 7 shows
threshold value of 256 bytes. the various scenarios designed using varying fragmentation
thresholds. In this section average media access delay (sec)
shown in Figure 16 and average throughput shown in Figure 17
has been investigated .
COMO2O1-7
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
REFERENCES
[1] Muhammad Ahsan, “Access Methods (Contention & Token
Passing),”
http://www.suite101.com/article.cfm/laning_retired/44958
[2] Todd Lammle, “CCNATM Cisco Certified Network
Associate Study Guide, 5th Edition,2006: BPB
Publications”.
[3] Behrouz A Forouzan, “Data Communications and
Networking, 4th Edition, 2007: Tata McGraw Hill”.
[4] Sameh H. Ghwanmeh, “Wireless network performance
Figure 17 Throughput (bits/sec) for varying fragmentation threshold (Global) optimization using Opnet Modeller”, Information
Longer frames tend to have higher error rates therefore Technology Journal 5(1): pp 18-24, 2006.
the transmission of long frames increases both the number of [5] ISO/IEC 8802-11: 1999, “Wireless LAN Medium Access
required retransmissions and the amount of data that is dropped. Control (MAC) and physical Layer (PHY) specifications,
To combat this inefficiency the 802.11 standards specify a Standard reaffirmed in June 2003", 1999.
mechanism for fragmenting frames when their length is above a [6] Bradley Mitchell,”Wireless Standards - 802.11b 802.11a
user-defined threshold. Fragmentation is allowed in the IEEE 802.11g and 802.11n: The 802.11 family explained”,
802.11b standard at the MAC layer; so that the fragment size available at
could vary from 100 to 2350 bytes. When a frame is fragmented http://compnetworking.about.com/cs/wireless80211/a/aa80
each segment is transmitted as if it were a separate frame, while 211standard_2.htm
the fragments are identified at the destination using various [7] Sarah Shaban, Dr. Hesham M.El Badawy, Prof. Dr. Attallah
fields in the packet header. Packet fragmentation allows large Hashad, “ Performance Evaluation of the IEEE 802.11
packets to be broken into smaller units when sent over the air, Wireless LAN Standards”, in the proceedings of the World
which is useful in congested environments or when interference Congress on Engineering-2008 , vol. I, July 2-4, 2008.
is a factor, since larger packets have a better chance of being [8] A. Goldsmith, “Wireless Communications, 1st Edition,
corrupted. This technique reduces the need for retransmission in August 2005: Cambridge University Press”.
many cases and thus improves overall wireless network [9] Manju Sharma and Manoj Kumar, “ HTTP and FTP
performance. The MAC layer is responsible for reassembling Statistics for wireless and wire-line network with and
fragments received, rendering the process transparent to higher without Redundancy based on OPNET” in the proceeding
level protocols. This fragmentation is implemented using the of National Conference on Optical and Wireless
built-in SAR (segmentation and reassembly) package in OPNET Communication (NCOW-2008)”, pp:70-75,September 27-
[2]. But fragmentation again increases the overall delay due to 28,2008.
the reassembling of multiple fragments at the destination. It also [10] Dr. Mohammad Hussain Ali and Manal Kadhim Odah,
increases the channel reservation time per data frame exchange. “Simulation Study of802.11b DCF using OPNET
Fragmentation threshold is thus an important parameter that Simulator”, Eng. & Tech. Journal,
affects WLAN performance. It is used to improve the WLAN vol.27,No6,2009,pp:1108-1117,2009.
performance when the media error rate is high. Packet [11] Hafiz M. Asif, Md. Golam Kaosar, “Performance
fragmentation for every data rate must be used to achieve Comparison of IP, ATM and MPLS Based Network Cores
throughput optimization (Figure 17). Using OPNET,” in 1st IEEE International Conference on
Industrial & Information Systems (ICIIS 2006), 8-11
III. CONCLUSION August, 2006, Sri Lanka.
In this paper, performance analysis of wireless computer [12] Manju Sharma, Manoj Kumar and Ajay K Sharma, "HTTP
networks has been done for improving the performance of and FTP Statistics for Wireless and Wire-line Network with
wireless LAN. The investigations reveal that the performance of and without Load Balance Based on OPNET",Published in
Wireless LAN can be improved by fine tuning and properly International Journal of Information and Systems Sciences,
choosing the WLAN parameters. For the tested simulation Volume 5,Number 1, pp. 112-125, 2009.
scenarios the performance is observed to be better with infra-red [13] Bradley Mitchell,” Wired vs. Wireless Networking:
type physical characteristics, higher buffer size (1024Kb), and at Which is better - wired or wireless?” available at
11 Mbps data rate. Increased throughput was also observed with http://compnetworking.about.com/cs/homenetworking/a/ho
mewiredless_3.htm
COMO2O1-8
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
[14] Arti Sood, “Network Design by Using Opnet™ IT Guru [22] Dr. Reinhard Kuch, “Studienbrief 2: Simulation of
Academic Edition Software", Rivier Academic Journal, networks using OPNET ITGURU Academic Edition v
2007, Vol. 3 (1). 9.1”, version: 1.0, 26-04-2009.
[15] Dibyendu Shekhar, Hua Qin, Shivkumar Kalyanaraman, [23] Marco Galeazzi, “Lezione 11: OPNET IT Guru Academic
Kalyan Kidambi, “Performance Optimization of TCP/IP Edition 9.1 A.A. 2007-2008 - OPNET ITGAE 9.1 (1/5)”
over Asymmetric Wired and Wireless Links” in the available at
proceeding of conference on Next Generation Wireless http://www.tlc.deit.univpm.it/sistemi/AA2007_08/Lezione1
Networks: Technologies, Protocols, Services and 1.pdf
Applications (EW-2002),February 25-28, 2002 – Florence, [24] “Opnet_Modeler_Manual”, available at
Italy,2002. http://www.opnet.com
[16] Guiomar Corral, Agustin Zaballos, Cesc Canet, “Proposal [25] T. Velmurugan, Himanshu Chandra and S. Balaji, “
of new challenge labs for the OPNET IT Guru Academic Comparison of Queuing disciplines for Differentiated
Edition”, available at Services using OPNET”, IEEE, ARTComm.2009.128, pp.
www.salle.url.edu/~zaballos/opnet/OPNET2004a.pdf 744-746,2009.Y
[17] Born Schilling, “Qualitative Comparison of Network
Simulation Tools,” 2005, www.ipvs.uni-
stuttgart.de/abteilungen/vs/lehre/.../CUBUS.../schilling.pdf
[18] Xinjie Chang, “Network Simulations with OPNET,”
Proceedings of Winter Simulation Conference, 1999, pp.
307-314
[19] “Introduction to OPNET Simulator” available at
http://bolero.ics.uci.edu/~ypan/OPNET/Introduction to
OPNET simulator.pdf
[20] “OPNET IT Guru Academic Edition” available at
http://www.opnet.com/university_program/itguru_academic
_edition
[21] Ranjan Kaparti, “OPNET IT GURU: A Tool for
Networking Education”
http://www.opnet.com/university_program/teaching_with_o
pnet/textbooks_and_materials/materials/ITGAE_Tool_Ntw
rk_Ed.pdf.
COMO2O1-9
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Abstract—A number of application level multicast protocols Mobile ad hoc network (MANET) is based on rapid
have been proposed for core selection and core migration in deployment of independent mobile users. The routers are
Mobile Ad Hoc Networks. Core migration is necessary to free to move randomly and organize themselves arbitrarily
minimize any disruptions on the transmission of data due to thus; the network‟s wireless topology may change rapidly
the changes in tree structure and to achieve improvement in and unpredictably [1]. Such a network may operate in a
the delivery of media streams in multicast group. In this stand-alone fashion, or may be connected to the Internet.
paper, core migration is performed on the varying size Multihop, mobility, large network size combined with
network graph model. The nodes within this ad hoc
device heterogeneity, bandwidth, and battery power
arrangement take on the values of the edge (We) and node (W)
randomly. With the varying numbers of iterations done on the constraints [19] make the design of adequate routing
core migration algorithm, every node and every edge picks up protocols a major challenge. Multicast promises efficient
different random values. The core migration in wireless ad use of network bandwidth for multiparty communication by
hoc network has been achieved by comparing the experimental allowing point-to-multipoint communication.
results based on the node weight and the edge weight. Recent [5] multicast protocols such as, Protocol
Independent Multicast (PIM) and Core Based Trees (CBT)
are based on the notion of group-shared trees. Multicasting
Keywords—Core-based multicast, mobile ad hoc network, core
routing protocols generally build trees to deliver messages
based tree, dynamic multicast membership, and core
migration to a multicast group. Delivering the information only
through edges belonging to the tree generates an efficient
I. INTRODUCTION form of multicast communication minimizing the amount of
In the past few years, a new wireless architecture has the network resources as required with the unicast routing.
been introduced that do not rely on any fixed infrastructure. In PIM, a group-shared tree is rooted at a rendezvous point
In this architecture, all nodes are mobile and no node plays (RP) similar to the CBT, which is rooted at the core node.
any special role. In fact, nodes reach other nodes they need In a multicast tree, the core node maintains the flow of
to communicate with using their neighbors as shown in traffic in the network. The [14] selection of this core node
figure1. Nodes that are close to each other discover their influences the shape of the multicast tree influencing the
neighbors. When a node needs to communicate with performance of the multicast tree with respect to the amount
another node, it sends the traffic to its neighbors and these of the delay experienced by leaves of the tree. Due to the
neighbors pass it along towards their neighbors and so on. dynamically varying nature of the network topology, nodes
This repeats until the destination of the traffic is reached need to be migrated to the one that satisfies the specified
[6]. Such architecture requires that every node in the QoS metrics. Migration of the core node takes place after
network play the role of a router by being able to determine selecting a core node from the set of candidate core nodes
the paths that packets need to take in order to reach their that may be either the neighboring nodes or the nodes
destinations. adjacent to the neighboring nodes of the previous core
node.
In this paper [16] investigate the problem of finding a
best node in terms of maximum edge weight (distance
between two communicating nodes) and weight of
individual node. This implies the movement of the core
node on a hop-by-hop basis to reach an optimal location.
The rest of this paper is organized as follows: Section 2
discusses the basics of wireless ad-hoc network. Section 3
covers the need and the techniques of the core migration
followed by the results and conclusion the next two
Fig. 1. Ad hoc network (client to client) sections.
COMO202-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
II. WIRELESS AD HOC NETWORKS list. Furthermore, it forwards the QUIT-REQUEST to its
parent node if its children list becomes empty and it itself is
A. Network Model not a multicast group node.
A mobile ad hoc network can be modeled as a network The CBT protocol provides mechanisms to deal with
consisting of n identical mobile hosts (nodes). These mobile link/node failures and the core node failure. Link failures
hosts employ a packet radio network to communicate with are detected by the exchange of periodic keep-alive
each other. Multiple nodes falling in the radio coverage area messages between the neighboring nodes. In case a node
of another node are said to be neighbors of this node and does not receive a certain fixed number of consecutive
they can simultaneously receive a message transmitted from keep-alive messages from its parent node, it assumes that
that node. The topology of the mobile ad hoc network is the link to its parent node has failed. In case a non-core
dynamic. Such a network can be represented by an node detects that its parent node or link to its parent node
undirected graph G = (V, E, W, We), where V is the set of has failed, it has two options for failure recovery: it can
nodes represented V = Nx and E is the set of logical links attempt to rejoin the multicast tree by sending a JOIN-
between neighboring nodes, W is the weight of each REQUEST towards the core node and it can send a
corresponding node in a set of V nodes, and We represents FLUSH-TREE message downstream, thus allowing each
the weight of the edge connecting the two nodes. It is node in the sub-tree rooted at it to independently attempt to
assumed that nodes leave and join the multicast group reattach to the multicast tree [63]. The first option results in
arbitrarily. the formation of loops in the multicast tree and, hence, the
CBT protocol provides a loop detection mechanism.
B. Core Based Tree Core-based multicast routing provides a good mechanism
The core-based tree protocol is a network layer multicast for scalable multicasting. However, migration of a core to
routing protocol that builds and maintains a shared delivery an “optimal” location is an essential task. In static networks,
tree for the multicast routing. Core-based tree is suited to core nodes may be pre-assigned and manually configured.
inter- and intra-domain multicast routing in the Internet. To However, in mobile ad hoc networks this may not be
establish the path between the sender and the receiver, core- optimal since the topology often changes. Dynamic
based tree may use a separate multicast routing table, or it selection and hence the migration of center nodes is
may use unicast routing. The core migration protocol important for good performance. A good core node may be
followed by the core selection protocol is designed to one, which is at the center of the portion of a network which
construct core- based tree. The purpose of the core spans all the multicast of the multicast group, which is the
migration algorithm is to locate that network node or a member of the multicast group, and which is fairly stable
router whose use as the core of the multicast group results [18]. This core migration protocol is based on the edge
in the best-multicast tree with respect to the desired QoS weight of current multicast tree. Core-based multicast tree,
specified by multicast application. The core-based tree is as discussed about, have been effectively used in solving
the shortest path tree rooted at some core node. The core the multicast routing problem in mobile ad hoc networks. In
node is also referred to as center node or a Rendezvous this work, this technique has been used to find a core node
point [16], [22]. The core-based tree protocol was in which qualitative factor are of major significance with
developed for Internet [18]. It uses a shared multicast tree the objective to extract and exploit the goodness of routing
rooted at core node, Rendezvous point or center node. It and multicast strategies to reach at the best possible route in
uses a receiver-based tree construction. A node interested in mobile ad hoc networks. Performance of the core migration
joining the multicast tree associated with a certain multicast method is generally measured using delay as a metric that is
group sends a JOIN-REQUEST towards the core node of the distance of the child (downstream) node from its parent
that group. The JOIN-REQUEST is routed towards the core (upstream) node in the entire graph.
node using the underlying unicast routing protocol. An
intermediate node receiving a JOIN-REQUEST simply C. Dynamic Multicast Membership
forwards it to the next node in the route towards the core Multicast group membership is dynamic, nodes are able
node unless it happens to be an established node in the to join and leave the group at any time. One of the multicast
multicast tree in which case the node responds by sending group members randomly assigns the responsibility of a
the JOIN-ACK to the node, which initiated the JOIN- group leader viz. a core to the first node of a multicast
REQUEST. A node traversed by JOIN-ACK record the group which then starts broadcasting RREQs. Other group
parent (upstream) node and the child (downstream) node. In members on receiving RREQs unicasts RREPs along the
effect, the JOIN-ACK grafts a tree branch from the node, shortest path to the core. Therefore the nodes are joined to
which responded with the JOIN-ACK, and the node, which the tree when their corresponding RREPs are accepted
initiated the corresponding JOIN-REQUEST. A node either by the existing members or by the core node.
wishing to leave the multicast group simply sends a QUIT-
REQUEST towards the core node along its tree branch. A III. FUNDAMENTALS OF CORE MIGRATION
node receiving a QUIT-REAUEST deletes the node along
which it received the QUIT-REQUEST from the children
COMO202-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
A. Core Selection position equalizes the edge weight (delay) coming from all
Core selection is the problem of selecting the placement the subtree branches [13]. The core migration is based on
of a core or cores in the domain for the purpose of the core-based multicast routing. This means that, if the
improving the performance of the tree(s) constructed performance metrics such as edge weight and node weight
around these core(s). The core selection protocol is of an adjacent core node is superior to the previously
designed to construct core-based trees using one of the considered node, then that adjacent core will be addressed
several different QoS metrics. The core selection algorithm as the migrated core and the members of the multicast
tries to find a router whose use as the core of the multicast group will then use this migrated core node as their new
group results in an optimal multicast tree with respect to a core node [22]. Typical core-based routing selects one
designed performance metric. The selection of the core is router as a core for a multicast group and builds a single
based on the explicit as well as implicit core selection multicast tree rooted at the core to deliver data to the entire
algorithms where all network nodes of the multicast group multicast group [24].
compute their weight functions and exchange weights In this core migration approach, the core determines the
among themselves so as to select the node with minimum edge weight over the several edge weights to find a
weight as the core for the latter case, and for the former neighbor experiencing the highest branch weight. The
case, the core monitors the delay by calculating the time neighbor with the highest branch delay or weight is selected
difference relative to the itself between the transmission of by the core as the new core if and only if the final migration
a packet and their corresponding acknowledgements [28]. to this node will not increase the delay on other branches
Core-based multicast routing scheme allows the migration [14].
of the core node by selecting the candidate cores (the nodes 1) TEMP_CORE Selection Method: The selection of the
adjacent to the core node). The selection of the candidate intermediate node is termed the TEMP_CORE, which has
node is based on the heuristic of the comparison between been selected for temporary basis until you find another
the current core and the set of the adjacent nodes. The node optimal node based on the edge and the node weight. This
with the significant better performance (in terms of the node gets selected every time the algorithm is run for
distance between the adjacent nodes ad their individual finding an optimal position. For the node weight algorithm,
weight) than the current core is elected as the new core and the lowercase node with largest path delay is selected and
finally migrated. The migration of the core takes place for for the edge weight algorithm, that particular node is
the node having largest path delay and minimum node considered as the intermediate node having the minimum
weight, resulting into the optimal core selection and weight value as mentioned by the table I. The on-route node
migration. In a typical ad hoc environment, network hosts with the largest path delay accounts for the largest amount
work in a group to carry out a given task. Multicasting is of time required for the transmission of the queries sent and
used for group-oriented communications. their corresponding replies received. Then it is required to
These core selection techniques given in [23] used Core- migrate to this on-route node after comparing the weight
Based tree having a single node, or router, which acts as the parameters of the current core with all its children nodes.
core of the tree, which enhances the scalability of the The migration to the on-route node with largest path delay
multicast algorithm, also the tree creation was receiver and minimum node weight is desirable.
based. The core selection protocol was designed to
construct core-based tree so that the core located for the C. Reasons for Migrating the Core Node
multicast group resulted in the best-multicast tree with Due to unconstrained movement of nodes in mobile ad
respect to the desired QoS specified by the multicast hoc networks, the topology of the network keeps changing.
application. Core Migration can be invoked when the quality of the tree
Performance of the core selection method was measured degrades due to membership dynamics or w tree followed
using Bandwidth i.e. the sum of all weights from the by addition of the node to the tree. Further, new nodes may
join the multicast group and hence the tree [18]. Hence, a
B. Core Migration
node which is a good core node for the multicast tree at a
Core migration occurs after selecting a new core for a given time may not remain good due to these changes.
multicast group. During the lifetime of a multicast Hence, a new core node must be periodically chosen.
application, the members of a group may change, and the However, changing the core node involves informing all the
resources in the network may fluctuate. The motive of the other nodes of the new core node as well as modifying the
ore migration is to identify a new core node for the group multicast tree so that the new core becomes the root of the
whose corresponding multicast tree, determined by the multicast tree. The core migration in the multicast tree of
current set of group members and present network status, mobile ad-hoc network is invoked due to [1]:
will likely result in significantly better multicast
performance than that of the current core [6]. For a fixed 1) Recovery from Core Failure: In core-based
network topology, the core migrates to an optimal position multicasting, core is a single point of failure. If the core
on a hop-by-hop basis as opposed to the dynamic network fails, there is a large amount of packet loss, as many of the
topology where the migration of the core to an optimal receivers cannot receive the data sent by the senders.
COMO202-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Therefore, recovery from both core failure and node/link 2. Then it waits for (Edge_Reply) message from all
failure are important issues. its ORNs.
3. On Route Node List is the union of all the received
2) Core Degeneration & Migration: Due to the dynamic
edge weights (here the delays are entered
nature of multicast group, core „degenerates‟ with time.
randomly)
That is, the quality of the core and hence the multicast tree
ORNL = U (received delays from all ORNs)
will deteriorate with time. This means that the core
4. The node with the largest edge weight is
migration has to be invoked when the quality of the core
considered as the Temp_Core.
degenerates.
5. It then finds the edge with the maximum weight by
Core migration involves the following: (i) selecting a set of candidate nodes.
selecting a new core that can offer better performance than 6. If the selected candidate core has the maximum
the current core, (ii) constructing a multicast tree based on edge value then it will be considered as the new
the new core, and (iii) migrating the group members from core.
the current multicast tree to the new multicast tree (tree 7. Else pick the adjacent node with larger delay to be
migration). Thus, core migration plays a vital role in tree the next Temp_Core, until the node with larger
maintenance and core failure recovery. delay is found.
D. Parameterized Techniques 2) Edge Weight: An algorithm based on the path delay
When the membership migrates throughout the network, between the two nodes is as follows:
current routing technique involves selection of a “core” 1. Core multicast the (Node_Query) message to all its
router through which all multicast communication is routed On Route Nodes (viz. to children nodes, here core
[11]. Then the routing scheme is adapted efficiently to is the parent node for there ORNs)
support randomly varying groups by allowing the core node 2. Then it waits for (Node_Reply) message from all
to migrate. In order to evaluate the performance of a core its ORNs.
migration method, and therefore that of the resultant 3. On Route Node List is the union of all the received
multicast trees, the following techniques are generally weights (here the node values are entered
considered. randomly)
ORNL = U (received weights from all ORNs)
1) Node Weight: It indicates the unique value assigned to
every node in the network. The core node compares its 4. The node with the minimum node value is
weight with all the neighboring nodes. If the weight of any considered as the Temp_CORE.
5. It then finds the node with the minimum weight by
of the neighboring or the adjacent node is lower than the
selecting a set of candidate nodes
weight of the core node, then it will stop looking for
6. If its own weight is less than lowest adjacent node
another node and the node with the lower weight will be
considered as the new Core. The core migration is based on weight viz. Temp_Core weight < weight of the
the core-based multicast routing. This means that, if the lowest adjacent node
7. Then migrate the core (viz. now the Temp_Core
techniques such as path delay and weight of an adjacent
will be considered as the new core)
core node are superior to the previously considered node,
8. Else pick the adjacent node with smaller node
then that adjacent core will be addressed as the migrated
value to be the next Temp_Core, until the node
core and the members of the multicast group will then use
this migrated core node as their new core node. Simulations with less weight is found.
can be done using high level programming using C++ or
any another platform.
2) Edge Weight: It is the distance between the core and
the child nodes adjacent to it. The distance of the core from
its group members should be less for efficient routing. The
core monitors the delay of all its adjacent nodes on a hop-
by-hop basis. Delay represents the weight of the edge
(connecting the two nodes). Therefore, a node with the
largest delay is taken for the new core node.
E. Core Migration Algorithm
1) Node Weight: An algorithm based on the individual
weight of the node is as follows:
1. Core multicast the (Edge_Query) message to all its
On Route Nodes (viz. to children nodes, here core
is the parent node for there ORNs)
COMO202-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
IV. RESULTS
The qualitative analysis of the core migration protocol
for wireless ad hoc network has been carried out by
comparing the two different algorithms for core migration
such as: edge weight algorithm and the node weight
algorithm.
COMO202-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
1 2 23 4 29 5 31 8 32 TABLE IV
2 2 23 4 31 5 33 CORE MIGRATION USING 10 NODES FOR EDGE WEIGHT
3 5 12 9 13 10 35
4 5 5 6 18 8 30 10 40
5 5 18 8 30 7 33 10 40 TABLE V
V. CONCLUSIONS
Research in wireless networks is progressing very fast
and numerous researchers from various fields focusing to
develop some workable scheme to find the best core node
or migration of the core if necessary. The motive of this
paper was to migrate the core to an optimal position by
COMO202-6
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
considering the two different techniques (edge weight and [10] M. Fleury, A. C. Downton, and A. F. Clark, “On the
node weight) related to the core migration. These two performance and feasibility of multicast core selection
parameterized techniques take on the random values every heuristics,” IEEE Transactions on Parallel and Distributed
time the core migration algorithm is run. This paper Systems, vol. 11, no. 11, pp. 164+, 2000.
[11] M. J. Donahoo and Ellen W. Zegure, “Core Migration for
suggests the need and possible technique of core migration Dynamic Multicast Routing,” Proceedings of International
wherever applicable. The results for the core migration in Conference, Computer Communication and Networks, 1996.
wireless ad hoc networks obtained with edge weight were [12] Mario Gerla, Rajive Bagrodia, et.al, “A Performance
verified with that obtained with the node weight, and it was Comparison Study of Ad Hoc Wireless Multicast Protocols,”
concluded that the core migration achieved with the edge in Proceedings of 19th Annual Joint Conference of the IEEE
weight technique results into the optimal location of the Computer and Communications Societies (IEEE INFOCOM
core than that obtained from the network graph modeled 2000).
using node weight parameter. This was due the fact that the [13] Manish Kochhal, Sandeep Gupta et.al, “An Efficient Core
edge weight considered the distance between the parent and Migration Protocol for QoS in Mobile Ad Hoc Networks,”
IEEE International Performance Computing, and
the child node and hence maintained a path delay between Communications Conference, pp.387-391, 2002.
the nodes which was not achieved with the node weight [14] Manish Kochhal, Sandeep Gupta et.al, “QoS-Aware Core
algorithm. But on the whole, core migration achieved with Migration for Efficient Multicast in Mobile Ad Hoc
node value parameter was faster compared to the edge Networks,” IEEE International Performance Computing, and
weight parameter. Communications Conference, October 2002.
[15] Nen-Chung Wang and Si-Ming Wang, “An Efficient Location-
VI. FUTURE WORK Aided Protocol for Mobile Ad Hoc Networks,” Proceedings
Further some other parameters except delay and node of 11th International Conference on Parallel and Distributed
systems, Vol.1, pp. 335-341, July 2005.
weight such as hop count, total path cost, throughput, and
[16] Rahul Malhotra and Savina Bansal, “Investigation of Core
bandwidth need to be considered in future when migrating Selection techniques in multicasting,” M.Tech thesis, 2008.
the core to a best optimal position. [17] Roger Zimmermann, Rahul Hampole, and Beomjoo Seo,
“Distributed Core Migartion in Multicast Peer-to-Peer
REFERENCES Streaming,” in Proceedings of International Conference on
[1] A. Chakrabarti and G. Manimaran, “A case for Scalable Parallel and Distributed processing techniques and
Multicast Tree Migration,” Global Telecommunications Applications (PDPTA 2004), June 21-24,2004.
Conference (GLOBECOM 2001), IEEE, Vol.3, pp. 2026 – [18] S.K.S.Gupta, P.K.Srimani et.al, “Adaptive Core Selection and
2030, August 2001. Migration method for Multicast Routing in Mobile Ad Hoc
[2] A. Chakrabarti and G. Manimaran, “A case for tree migration Networks,” IEEE Transactions on parallel and Distributed
and integrated tree maintenance in QoS multicasting,” Systems, Vol.14, No.1, January 2003.
Computer Communications, Volume 26, Issue 9, No.2, pp. [19] Stephen Mueller, Rose P. Tsang, and Dipak Ghosal,
1007-1017, June 2003. “Multipath Routing in Mobile Ad Hoc Networks: Issues and
[3] A.Dvir and M.Segal, “Placing and Maintaining a Core Node Challenges,” In Performance Tools and Applications to
in wireless Ad Hoc Networks,” in Wireless Communications Networked Systems, volume 2965 of LNCS 2004.
and Mobile Computing, Vol.00, pp.1-14, 2008. [20] Sudarshan Vasudevan, Jim Kurose and Don Towsley, “Design
[4] Ayse Karaman and Hossam Hassanein, “Core-Based and Analysis of a Leader Election algorithm for Mobile Ad
Approach in Multicast Routing Protocols,” 2003. Hoc Networks,” Proceedings of 12th IEEE International
[5] C.V.Ravishankar et.al, “Distributed Center- conference on Network protocols (ICNP), pp. 350-360,
Location Algorithms,” IEEE Journal on Selected October 2004.
Areas in Communications, Vol.15, No.3, pp.293-303, [21] Ting-Yuan Wang, Lih-Chyau Wuu, and Shing-Tsaan Huang,
April 1997. “A Scalable Core Migration Protocol for Dynamic Multicast
[6] Eric Fluery, Philip K. McKinley, and Yih Huang, “LCM: A Tree,” Journal of Information Science and Engineering,
Multicast Core Management Protocol for Link-State Routing Vol.19, pp.479-501, 2003.
Networks,” in Proceedings of IEEE International Conference [22] Deering, S., Estrin, D. L., Farinacci, D., Jacobson, V., Liu, C.
on Communications, Vol.2, pp.1197-1201, June 1998. and Wei, L., (1996), “The PIM Architecture for Wide-Area
[7] Hwa-Chun Lin, Zhe-Hong Lin, “Selection of Candidate Multicast Routing,” IEEE/ACM Transactions on Networking,
Cores for Core-Based Multicast Routing Architectures,” in 4(2), pp.153-162.
Proceedings of IEEE International Conference on [23] Gerla, M., Chiang, C. C., and Zhang, L., (1999), “Tree
Communications, pp.2662-2666, 2002. Multicast strategies in mobile, multihop wireless networks,”
[8] Jean-Francois Macq and Michel X. Goemans, “Trade-offs on In Mobile Networks and Applications, 4 (3), pp. 193-207.
the Location of the Core Node in a Network,” Wiley [24] Puthhividhya, W., Tavanapong, W., Tran, M., and Wong, J.,
InterScience Vol. 44, Issue: 3 (October 2004) pp. 179 – 186, (2004), “Distributed Core Selection with QoS Support,” In
2004. IEEE International Conference on Commumnications, 4, pp.
2132-2137.
[9] Kenneth L. Calvert, Ellen Witte Zegura, and Michael J.
Donahoo, “Core selection methods for multicast routing,” in
Proceedings of the 4th International Conference on Computer
Communications and Networks (ICCCN ‟95), Las Vegas,
Nevada, pp. 638+, September 20–23, 1995
COMO202-7
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Abstract—In interleave division multiple access (IDMA), version of CDMA, which inherits many advantages from
interleavers are used for the user separation. This paper CDMA such as dynamic channel sharing, mitigation of
presents the performance comparison between random and cross-cell interferences, asynchronous transmission, ease of
tree based interleaver based IDMA system, in fading cell planning, and robustness against fading. It also allows a
environment. We also know that Maximal Ratio Combining low complexity multiple user detection (MUD) techniques
(MRC) diversity approach is popularly used to mitigate the [7] (CBC detection) applicable to systems with large
fading problem in any wireless system so here MRRC scheme numbers of users in multi-path channels.
is used with both RI & TBI based system to mitigate fading.
The approach involves receiver diversity analysis as 1:2 The objective of this paper is to use MRC diversity in
antenna systems in fading environment; no. of antennas can be IDMA communication system to reduce the effect of fading.
increased to improve diversity order. The study of transmit as well as receiver diversity is taken
separately, because both have their own application area [2].
The paper is organized as follows. Concept of IDMA is
introduced in section II. Section III deals with classical
Keywords- Random Interleaver,Tree Based Interleaver, MRRC diversity approach used with IDMA. In section IV
MRC diversity, IDMA
transmit diversity is discussed with IDMA. Performance
I. INTRODUCTION analysis is provided in section V. Finally conclusions are
presented in section VI.
The goal for the next generation mobile communications
system is to seamlessly provide a wide variety of II. IDMA SCHEME
communication services to anybody, anywhere, anytime
such as high voice quality, higher data rates etc. The A. IDMA Mechanism
technology needed to tackle the challenges to make these The performance of conventional code-division multiple-
services available is popularly known as the Third access (CDMA) systems [1] is mainly limited by multiple
Generation (3G) Cellular Systems using multiuser detection access interference (MAI), as well as intersymbol
[1]. The fundamental phenomenon which makes reliable interference (ISI). Also, the complexity of CDMA multi-user
communication difficult is time varying multipath fading, detection has always been a serious problem for researchers
which is major impairment in any wireless communication all over the world. The problem can be visualized from the
system. The performance improvement is very difficult in angle of computational cost as well complexity of multi-user
such situation detection algorithms in CDMA systems. The use of user-
specific signature sequences is a characteristic feature for a
Theoretically, improvement in signal to noise ratio may be conventional CDMA system. The possibility of employing
achieved by providing higher transmit power or additional interleaving for user separation in CDMA systems is briefly
bandwidth which are not feasible solution as they are inducted in [1] but the receiver complexity is considered as a
contrary to the requirements of next generation wireless main problem. In interleave-division multiple-access
communication [2]. On the other hand, the problem of fading (IDMA) scheme, users are distinguished by user specific
may be handled with suitable diversity technique without chip-level interleavers instead of signatures as in a
expanding communication resources easily. conventional CDMA system. The scheme considered is a
In most wireless channels, antenna diversity is a special case of CDMA in which bandwidth expansion is
practical, effective and widely used technique for reducing entirely performed by low-rate coding. This scheme allows a
the effect of multipath fading. The maximal ratio combining low complexity multiple user detection techniques applicable
(MRC) diversity technique, is implemented with interleave- to systems with large numbers of users in multipath channels
division multiple-access (IDMA) scheme, as MRC is in addition to other advantages.
performed well in comparison with selection or equal gain
combining.[3]-[5]. The IDMA scheme is known as advanced
COMO203-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
In receiver section, after chip matched filtering, the is the distortion in r( j) with respect to user-k. k ( j ) is
received signal form the K users can be written as
the distortion (including interference-plus-noise) in received
K signal with respect to user-k.
r( j) hk xk ( j ) n( j ), j 1, 2,.......J . (1)
k 1 A brief description of CBC algorithm [1] used in IDMA,
th
has been presented in [3]. The operations of ESEB and APP
where h k is the channel coefficient for k user and decoding are carried out user-by-user.
{ n( j ) } are the samples of an additive white Gaussian The outputs of the ESEB as extrinsic log-likelihood
noise (AWGN) process with mean as zero and variance σ 2 ratios (LLRs) is given as,
=N0 / 2. An assumption is made that {h k} are known priori at r ( j ) E (r ( j )) hk E ( xk ( j ))
the receiver. eESEB ( xk ( j )) 2hk . 2
Var (rj ) hk Var ( xk ( j ))
The receiver consists of a elementary signal estimator
block (ESEB) and a bank of K single user a posteriori The LLR output of SDEC is given as,
probability (APP) decoders (SDECs), operating in an
iterative manner. The modulation technique used for S
simulation is binary phase shift keying (BPSK) signaling. eSDEC ( xk ( ( j ))) eESEB ( xk ( ( j ))) j 1,..., S
The outputs of the ESEB and SDECs are extrinsic log- j 1
likelihood ratios (LLRs) about {x k} defined as Now, these steps are repeated depending on no. of
iterations and users.
COMO203-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
4
=h0 R0 + h1 R1 (7)
Memory Requirement of Interleaver
K
With Random Interleaver
3.5
With Tree Based Interleaver
Now this output of maximal ratio combiner can fed to the 3
detector for the proper estimation of transmitted signal xk.
2.5
1.5
0.5
0
0 10 20 30 40 50 60 70 80 90 100
User Number
COMO203-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
extremely less than that required for random interleaver the reference of computational complexity and memory
generation method [3] and is independent of user count. requirement it takes edge on random interleaver. It can also
conclude that IDMA with suitable diversity technique can
generate fruitful results in the area of wireless
communication. study can also enhanced to Multiple input
and multiple output (MIMO) antenna system to improve the
diversity order and hence the performance of IDMA
Figure 4 demonstrates the performance of IDMA scheme communication system.
with using random interleaver and tree based interleaver.
Here we refer the channel as slow fading Rayleigh channel. REFERENCES
The interleavers used in simulations are random interleaver [1] M. Moher and P. Guinand, “An iterative algorithm for asynchronous
[3] and tree based interleaver [15]. The block length is 200 coded multi-user detection,” IEEE Commun. Lett., vol. 2, pp. 229–
231, Aug. 1998.
with spread length to be 16. The iterations selected for
[2] S.M.Alamouti. “A Simple transmitter diversity scheme for wireless
simulation in receiver is 15. The simulations have been communications” IEEE J. Select. Areas Commun., vol.16 , pp. 1451-
performed for one transmitter and two receiver arrangement 1458, Oct.1998.
in the case of receiver diversity. [3] Li ping, Lihai Liu, Keying Wu and W. K. Leung, “Interleave-
Division Multiple-Access,” IEEE Trans. Wireless Commun., vol. 5,
The BER performance shows that although RI based IDMA pp. 938-947, April 2006.
performs better but TBI based system also performs good for [4] H. Wu, L. Ping and A. Perotti, “User-specific chip-level interleaver
larger value of Eb/N0 design for IDMA systems,” IEEE Elec. Lett., vol. 42, no. 4, Feb.
2006.
[5] Li Ping and L. Liu, “Analysis and design of IDMA systems based on
SNR evolution and power allocation,” in Proc. VTC‟2004-Fall, Los
Angles, CA, Sept. 2004
[6] S. Brück, U. Sorger, S. Gligorevic, and N. Stolte, “Interleaving for
-1
10 outer convolutional codes in DS-CDMA Systems,” IEEE Trans.
TBI with receive diversity.k=131072 Commun., vol. 48, pp. 1100–1107, July 2000.
TBI with receive diversity,k=65536
RI with receive diversity, k=65536 [7] Tarable, G. Montorsi, and S. Benedetto, “Analysis and design of
-2
10
RI with receive diversity, k=131072
interleavers for CDMA systems,” IEEE Commun. Lett., vol. 5, pp.
420–422, Oct. 2001.
[8] F. N. Brannstrom, T. M. Aulin, and L. K. Rasmussen, “Iterative
-3
multi-user detection of trellis code multiple access using a posteriori
10
probabilities,” in Proc. ICC 2001, Finland, June 2001, pp. 11–15.
Bit Error Rate
COMO203-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO205-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
can speed up the process of vulnerability mitigation overlay for streaming applications should not only
as similar vulnerabilities may require similar accommodate global design goals such as scalability
solutions. In other words, in order to take any action, and resilience but also satisfy the local design goal of
a level of priority can be determined based on known maximizing delivered quality to individual peers. The
data from other vulnerabilities. For example, a low commonly have the two drawbacks as follows. First,
priority level can be assigned to a newly found the systems have static overlay networks [5] consisted
vulnerability if it is proved that it is similar to a of reliable nodes under the administrative control, or
known vulnerability that has a low impact, while the assume that a spanning tree of entire node is known
basic idea is simple, a practical system design is not beforehand. Clearly, this is not feasible when a
easy. P2P reputation system are consists of three system involves an enormous number of nodes, which
functional components collecting information on peer join and leave the overlay network dynamically. In
behavior, scoring and ranking peers, and responding extreme case applications, nodes systems may be
based on peers’ scores. organized only by client nodes without any
All these components are nontrivial, especially specialized node. Second, a node keeps a large
given the following consideration [4]: amount of routing states and its control message
Scalability: A large P2P network may have overhead is huge. This is because every node can be
hundreds of thousands of peers. For example, Skype an intermediate router on the paths of an event
has several million online users. A reputation system dissemination tree. A distributed sender/receiver
should be highly scalable in terms of peer number. needs a structured overlay network, but it must be
Adaption to peer dynamics: Peers may join or designed with great care since the underlying peer-to-
leave at any time. If reputation information is peer overlay substrate has a significant effect on the
maintained at peers, peer leaving may lead to performance of the system.
information loss. A robust reputation system should To solve the above challenges, a new peer-to-peer
take peer dynamics into account. overlay; one important metric in evaluating
Security: Malicious peers may endeavor to break distributed content based systems is the percentage of
down the reputation system so that they can conduct false positives over the total number of messages. A
malicious actions without being detected. For false positive is defined as a message received by the
example, peers may purposefully leave and rejoin the node not interested in the message. This section
system with a new identity in order to shed any bad presents some preliminary evaluation of the security
reputation. protocol [9] using simulations.
There are two fundamental issues in P2P So peer-to-peer file-sharing are so dangerous?
reputation systems. Reputation estimation [10]: An There are a number of reasons:
estimation method describes how to generate peer Illegal, Unethical, and immoral--unfortunately the
reputation based on others’ feedback. These classify most common uses of P2P are illegal ones such as
existing estimation methods into three categories: copy-righted music, copyrighted software,
social network, probabilistic estimation, and game pornographic material, and even child pornography.
theoretic model. As many estimation methods rely on Viruses--when downloading a shared file, one big
specific feedback collection mechanisms the feedback danger is to download a computer virus or Trojan
collection mechanisms. The necessary networks are horse onto your computer. These viruses can cause all
not trivial. First, efficient data storage and retrieval is kinds of problems from erasing all files on the hard
always a challenging issue in P2P networks. Then drive to the automatic sending of pornographic e-
efficient retrieval becomes nontrivial. Peer dynamics mails to all of your friends in your mail directory.
bring more difficulty. Second, reputation data are Overlay Network Centralization--Although in their
highly security-sensitive. There are many other purest form peer-to-peer overlay networks are
important issues in P2P reputation systems; for supposed to be totally decentralized in practice this is
example, how to prevent targeted and adversarial not always true, and systems with various degrees of
attacks? How to interpret reputation? Interested centralization are encountered.
readers may refer to for a comprehensive overview of This typically renders them inherently un-scalable
P2P reputation issues to the end, each peer can and vulnerable to censorship, technical failure or
connect to the overlay topology [6] at multiple points malicious attack and peer to peer objective in figure 1
(i.e., receive content through multiple parent peers). shows.
Therefore, participating peers form an unstructured
overlay [2] that can constructing a peer-to-peer
COMO205-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO205-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Data delivery is the process of how the stream data is solutions that examine, as well as their relationship to
transported from the source to the destination. The the relevant non-functional characteristics, were
source, the destination and all the intermediate nodes assembled through a detailed analysis and current
in a streaming system participate in a topology that is peer-to-peer content distribution systems that are
constructed based on the specific system’s protocol in either deployed
a P2P streaming system. The most important are attributes of peer to peer
contents distribution systems namely [14, 15].
Security: Further analyzed in terms of:
Integrity and Authenticity: Safeguarding the
accuracy and completeness of data and processing
methods. Unauthorized entities cannot change data
adversaries cannot substitute a forged document for a
requested one.
Privacy and Confidentiality: Ensuring that data is
accessible only to those authorized to have access and
Figure 2: The Architecture of P2P Network that there is control over what data is collected, how
with security system it is used and how it is maintained.
Availability and Persistence: Ensuring that
Data Acquisition and Presentation: At the streaming authorized users have access to data and associated
source, the content is prepared for distribution. If the assets when required. For a peer-to-peer content
data was prerecorded and is available as files, it distribution system this often means always. This
categorizes this as on-demand streaming. Data property entails stability [7] in the presence of failure
Delivery: The transition of one or multiple copies of or changing node populations.
the content from a source node to a destination node Scalability: Maintaining the system's performance
is called a streaming session. A streaming session attributes independent of the number of nodes or
starts when a streaming request is made and ends documents in its network. A dramatic increase in the
when all associated destination nodes have received number of nodes or documents will have minimal
the last byte of the content. Depending on the number effect on performance and availability.
of source and destination nodes involved in a Performance: The time required for performing the
streaming session. Peer-to-Peer System Operation: operations allowed by the system, typically
From the perspective of a peer, the life-cycle of a P2P publication, searching and retrieval of documents.
streaming session can be decomposed into a series of Fairness: Ensuring that users offer and consume
four major processes: finding the service, searching resources in a fair and balanced manner. May replies
for specific content, joining or leaving the service, on accountability, reputation and resource trading
and failure recovery when there is an error. mechanisms.
Resource Management Capabilities: In their most
3.1 Analysis of Architecture basic form peer-to-peer content distribution systems
allow the publishing, searching and retrieval of
In this description and analysis of architecture, documents. More sophisticated systems may afford
systems and infrastructures that are based on peer-to- more advanced resource management capabilities
peer architectures and aim at either offering content such as editing or removal of documents,
distribution solutions or supporting content management of storage space and operations on
distribution related activities: metadata.
Our approach is based on: identifying the feature Semantic Grouping of Information: An area of
space of non-functional properties and characteristics research that has attracted considerable attention
of content distribution systems, recently is the semantic grouping and organization of
Determining the way in which the non-functional content and information in peer-to-peer networks.
properties depend on and can be affected by various Various grouping schemes are encountered, such as
design features, and semantic grouping based on the content itself,
Providing an account, analysis and evaluation of grouping based on locality or network distance,
the design features and solutions that have been grouping based on organization ties as well as others.
adopted by current peer-to-peer systems, as well as
their shortcomings, The various design features and
COMO205-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO205-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
to different situations (load balancing, failures, [8] Yongjin Choi and Daeyeon Park “Mirinae: A
network traffic, etc.). Network Security is well suited Peer-to-Peer Overlay Network for Large-Scale
to heterogeneous P2P networks. Our assumptions Content-Based Publish/Subscribe Systems” acm
show that Network Security efficiently handles 2005.
failures. It is highly stable and the resource oriented-
hierarchy provides an optimal use of the network. The [9] Laurent Massouli ´Erwan Le Merrer, Anne-
technique used to define and maintain the routing Marie Kermarrec and Ayalvadi Ganesh
tables allows the use of routing algorithms having a Microsoft Research, Cambridge, U.K. IRISA and
logarithmic complexity. FTR&D, Lannion, France INRIA/IRISA,
Rennes, France “Peer Counting and Sampling in
7. References Overlay Networks: Random Walk Methods”
2006.
[1] Wu Kehe, Zhang Tong, Li Wei, Ma Gang
“Security Model Based on Network Business [10] S. Zhou, M. Hogan, S. Ardon, M. Portman, T.
Security” IEEE 2009. Hu, K.Wongrujira, and A. Seneviratne “ALASA:
When Service Overlay Networks Meet Peer-to-
[2] Diego Doval and Donal O’Mahony Trinity Peer Networks” IEEE 2005.
College Dublin “Overlay Networks a Scalable
Alternative for P2P” IEEE 2003. [11] Mei Li Wang-Chien Lee Anand
Sivasubramaniam “Semantic Small World: An
[3] William Couch “Peer-to-Peer File-Sharing Overlay Network for Peer- to-Peer Search” IEEE
Networks: Security Risks” 2002 2004.
[4] Yan Chen, David Bindel, Randy H. Katz [12] Chunqiang Tang, Zhichen Xu, Sandhya
“Tomography-based Overlay Network Dwarkadas “Peer-to-Peer Information Retrieval
Monitoring” acm 2003. Using Self-Organizing Semantic Overlay
Networks” acm 2003
[5] Kenichi Watanabe and Yoshiona kajima
“Ranking Factors in Peer-to-Peer Overlay [13] www http://napster.com.
Networks” acm 2007.
[14] www http://vivisimo.com
[6] Arturo Crespo and Hector Garcia-Molina”
Semantic Overlay Networks for P2P Systems” [15] www.grokster.com
2004.
[7] Junghee Han, David Watson, and Farnam
Jahanian “Topology Aware Overlay Networks”
2004.
COMO205-6
ENERGY CONSERVATION THEORY IN CONTEXT AWARE COMPUTING
Umesh Chaudhary , Chandrika Prasad
Department of Computer Science, Banaras Hindu University
Varanasi-221005
umesh9794@gmail.com, chandrika.bhumca08@gmail.com
T1980s.
Wireless cellular system has been in use since
HE
Wireless system operates with the aid of
a centralized supporting structure such as an access
point. These access points assist the wireless users
to keep connected with the wireless system, when
they roam from one place to other. In wireless
system the device communicate via radio channel to
share resource and information between devices.
Due to presence of a fixed supporting structure,
limits the adaptability of wireless system, so this
generation of wireless system is required easy and
quick deployment of wireless network. Recent
Figure 1.1
advancement of wireless technologies like
In ad hoc network there can be node that will try to
Bluetooth [3], IEEE 802.11 [4] introduced a new
disrupt the proper functioning network. These nodes
type of wireless system known as Mobile ad-hoc
can be malicious or selfish [8]. They try to disrupt
network (MANETs) [1, 2, 5, 6], which operate in
network function by modifying packets, injecting
the absence of central access point. It provides high
packets or creating routing loops. So, security is an
mobility and device portability’s that enable to node
important task, because MANETs has
characteristics such as; dynamic topology,
COMO207-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 2
COMO207-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 3
COMO207-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 4
COMO207-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 5
COMO207-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 6
In the Network Initialization procedure, if a node is nodes (such as public key, address, timestamp t,
in IM and send the PackHello Packet. timestamp e, nonce and signature to all others node
that participate in the network)
Mode Packet Transition to
NodeName Publi Nonc Timestamp NextNodeName NumberofHop
exchanged Mode c key e s
procedure it again back to LM, and forward This mode is basically for monitoring the network
PackLazy packet to all neighbor nodes. topology when nodes leaving the network, joining
the network or changing its position with respect to
1. Start neighborhood. This mode also detect malicious and
2. Set Flag = 1 (i.e. If there is any activity of selfish node. If the node in mode found any activity
nieghbour, then) of neighbour node or receives any PackMalicious
Switch to MM Packet or PackSelfish Packet or PackForward
Else flag = 0 Packet it switches itself to MM. This is the main
End loop mode for ASRP, it provide high level of security.
3. It check the neighbour node, for malicious or This mode is also called the protector mode. These
selfish , if yes then send above condition can divided into two part; general
PackSelfish to all neighbour node condition and special condition.
4. If not then return to LM
5. Stop.
A. General Condition
I. Procedure for MM when node are in Lazy mode
and detect any activity of neighbor nodes
Figure 1.2 Transitions between Modes in LM II. Procedure for MM when nodes are in Lazy mode
II. Procedure and receive PackMalicious or PackSelfish or
The Lazy mode is called the default mode of PackForward to neighbour node.
the node.
As soon as the node finished the network B. Special Condition
initialization they switch from IM to LM. Nodes joining the network, nodes leaving the
If there is no PackForward packet to be network and nodes are changing its position within
forwarded they switch from PFM to LM. the network. These three special conditions are also
When the node in LM detect that there is handling in monitor mode by sending or receiving
activity of neighbor node it switch itself to MM PackUpdate packet. These three conditions are
If the PackForward packet comes from the explained below. In the case of new node joining
neighbor node then it switch them to PFM through the network, say node A want to join the network.
MM and after finishing packet forwarding There are two cases possible, first is the network
procedure it comeback to LM. has no any node initially and the second case is
If all the nodes want to change their public key node A join the network through nodes B1 to Bn. in
then they are also switch LM to IM. first case node A enter into network without any
In the nodes comes in the Lazy mode it send the condition. In second case if n=1, then it join through
PackLazy packet to all the neighbor node. B1 and if n=2, then it join through B1 and B2.
Monitor Mode
COMO207-7
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 8
a. The Procedure for node joining the network is The ASRP is proactive secure routing protocol, so
given below when MANET is established every source node
1. Start know the route to the all its possible destination
2. Flag Join = False node which are reachable from it. The source node
3. While ( Join = False) then prepare the packet and send it to first node
The node A checks that if at least one node out of which are on the route of destination node, then first
all the node through which it is node send to next in this way the data is reaches to
going to join the network. if there is one node, then the destination node. Now the packet that is sent,
Join = True consist two parts
Else Join = False PPart NPart
End Loop
4. The node A switch to IM and send the
PackHello packet to nodes out of B1 to Bn The first part i.e. Privacy part contain the data
which are in LM. which is to be transmitted is encrypted by source
5. The node which are in LM, out of B1 to Bn node by using public key of destination node. The
switch themselves to IM, when it receive NPart i.e. non-repudiation part contains the address
the PackHello packet from node A, then updates it of destination node, a nonce of destination node and
NodeInfo table and then it send the timestamp, and the address of all the node along
PackInitialized Packet to A. they transmitted. It also contains the address of
6. Node A match all the PackInitialized packet source and signature of source node which is
send by B1 to Bn if there is no change, encrypted by private key of source node. Every
then intermediate node along the decrypt the NPart of
Node A Update its NodeInfo table and send its Packet and verify its addresses, if itself destination
own information to all node within the node or not then forwards the packet to the next
network node in the path. This way packet reached to
All node in MANET Updates its own NodeInfo destination node. The NPart contains only the
table information related to route. This mode detect
Else send PackMalicious to all neighbours. malicious and selfish node by verifying NPart of
7. After finishing IM procedure they switch to LM packet. It also detect if link is broken to destination
8. The resident nodes again switch to LM. node or if destination not exists by generating
9. Stop. PackError packet. If any node behaves as a selfish it
b. Procedure when the node, say D changes its detect by resending packet again. If packet send by
position or detects the change in its neighborhood or source node to particular destination and the
node D leaving the network. intermediate node not find the particular destination
it send PackError packet to source node.
1. Start The data forwarding process is better understood by
2. Check any change in its neighbors, if node D using an example. Let us suppose the source node
leave the network, then want to send the data in PackForward packet has to
3. The neighbor node E of D send the destination node along the way to destination
PackUpdate to all its neighbor through the source A B C D destination.
4. If node D change its position, then D send
the PackUpdate to its neighbors
5. The neighbor node after receiving it switch
to MM, and updates their NodeInfo
table and send PackUpdate to all its own
neighbors
6. Stop.
Packet Forward Mode
COMO207-8
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 9
Source A {PPart, NPart} The ASRP is the protocol that works by enforcing
PPart =EPA (EPB (EPC (EPDestination (Data, Source, t, the cooperation between the nodes. That makes the
n, C), B,) A), Source) nodes are more responsible for their action and the
NPart = EPRSource (A, B, C, (Destination, t, n) probability of the non-intentional behavior of the
KSource- node also become less. The MANET does not pass
A B{PPart, NPart} the any infrastructure so the nodes have to cooperate
PPart = EPB (EPC (EPDestination (Data, Source, t, fully in order to enforce the operation on the nodes
n, C), B), A) in the correct way.
NPart = EPRA (B, C, Destination, t , n) KBB II. Robust Modular Implementation
C {PPart, NPart} There are four modes of the ASRP each mode has
PPart= EPC (EPDestination (Data, Source, t, n, C), B) its own modular implementation and independent
NPart = EPRB (C, Destination, t, n) KCC working. That paves the way for the modification of
EPDestination {PPart, NPart} the working of the modes in the future as the need
PPart= EPDestination (Data, Source, t, n, C) arises without modifying the working of the other
NPart = EPRC (Destination, t, n) modes. Moreover the modular approach is best for
the debugging also.
Simulation
Additional Security Features We are using turbo C, for the simulation, has been
The others security features enhanced the security of performed to simulate the transfer of data between
MANET, are given below. various nodes in initializing mode, lazy mode,
I. The key of the nodes have to be change regularly packet forward mode and monitor mode. The
after some specific time interval, so as keep the simulation has been performing on eight nodes.
MANET secure all time. Figure 1.4 Simulated MANET
The simulation program after taking the input
5 1 performs the simulation as mention below.
I. Network input by adding the node in form of
adjacency list
6 4 2 0 II. Simulate the packet transfer in the Initializing
Mode
III. Simulate the packet transfer in the Lazy Mode
7 3
IV. Simulate the packet transfer in the Packet
Forwarding Mode
Since the ASRP is a proactive secure routing
II. The dummy packet forwarding operation can be
protocol, so in every step it displays the status of
implemented by the intermediate node to detect the
tables of all the nodes. The simulation step along
node who misbehaves.
with screenshot has been given below. I. Firstly the
III. To detect malicious and selfish during the
simulation program takes the input, in terms of
network initializing process by verifying public key
asking for the number, name and name of neighbour
and signature send by all neighbour node to new
of nodes.
node.
IV. In case of special scenario of military operation,
like some nodes are to join the network latter on,
only those nodes join which has the MAC address
that has been already verify
Inbuilt Defense
The various inbuilt defense are in ASRP, to
securing MANET, these are:
I. Cooperation Enforcement
COMO207-9
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 10
Figure 1.7 Input Times in Millisecond Figure 1.9 Information about the route of all
destinations from a source
COMO207-10
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 11
b) Network after (establishing MANET) step II. Conclusion and Future Scope
Wireless mobile ad hoc networks present difficult
challenges to routing protocol designers. Mobility,
constrained bandwidth, and limited power cause
frequent topology changes. The very basic nature of
the mode of communication is the main concern
because anything that moves over the open air
medium is susceptible to be picked up by
unauthorized access. For any mission critical or
organizationally sensitive information, ad hoc
networks add an element of insecurity. In the existing
Figure 1.10 MANET after step II secure routing protocols most of the security attacks are
III. Packet Transfer in Lazy Mode possible with a compromised node. In this work we focus on
how to detect malicious and selfish node and to design and
implement a secure routing protocol.
In ASRP protocol we discuss various activity of node which
they are shown during the MANET operation and these
activities are grouped into modes along their working. We also
discussed the packets that are going to be exchanged in
different mode of nodes. The conclusion that comes are given
below.
I. The problems of malicious and selfish node are handling
simultaneously. We discussed Extended Public key
cryptography mechanism to handle the malicious and selfish
node during network operation. As the selfish node cannot
malicious at same time, but if nodes are not malicious then
Figure 1.11 Packet transfer during Lazy Mode
they may be malicious.
IV. Packet Forwarding Procedure
II. The protocol is handling the some special situation like
nodes joining the network, node leaving the network and nodes
are changing its position within the network. The monitor
mode of ASRP handles all three situations.
III. In ASRP, there are four modes, the IM corresponds to
network initialization phase, the LM corresponds to default
phase, PFM responsible for forwarding the packet form source
to destination and MM is the protector mode of the network.
IV. The protocol has to develop in the way so that the future
modifications are possible without changing overall protocol.
Future Scope
The proposed secure routing protocol, ASRP, is a
proactive routing protocol based on table driven
approach. For the future work we can use the hybrid
approach or reactive approach in ASRP to
implement a new secure routing protocol. We can
also add another mode or existing one can be
extended to handle some exceptional conditions.
The public key cryptography algorithm can also be
extended to securing MANET.
COMO207-12
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Abstract— This paper describes a study on RF numerical analysis and simulation in Matlab since this model
attenuation path loss behavior in suburban coverage shows good performance compared to other models. The
areas in Haryana. Hata model is the most popular model field measurement data was collected using suitable
that extensively used in Europe and North America. The equipment for outdoor measurements and is then compared
model developed by Y. Okumura and M. Hata and is with the simulation results.
based on measurements in urban and suburban areas at
Japan. There are practical problem faced in India if this II. THEORETICAL ANALYSIS
model is used for the prediction of the signal in suburban
areas of India. The main problem is that the path loss Hata model is the most popular model that extensively used
calculated with Hata model for less than 1km shows in Europe and North America. The model developed by Y.
large practical difference. This difference can be due to Okumura and M. Hata [3] and is based on measurements in
the empirical Okumura Hata model is not according to urban and suburban areas at Japan in 1968. Validity range of
the land profile of India, building and environmental the model is frequency fc between 150 MHz and 1500 MHz,
conditions are different from the basic inferences used in TX height hb between 30 and 200 m, RX height hm between
formulation of this model in Japan and the Vegetation 1 and 10 m and TX-RX distance r between 1 and 10 km.
and foliage losses etc. are different. An effort to improve Hata's Equation is classified into three models [4]:
the Hata empirical model for sub urban area is done in
this research paper. Rural: open space, no tall trees or building in path
Suburban area: Village Highway scattered with
Keywords- Path Loss; Okumura Hata model; Optimization; trees and house with some obstacles near the mobile
Suburban but not very congested.
Urban area: Built up city or large town with large
building and houses.
I. INTRODUCTION
Definition of parameters:
COMO209-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
(ii) Suburban Areas Path Loss, different BTS is available for post processing. The
2 measurement data is available in the log file [7].The details
fc
L(dB) = L(urban) -2 log 5 .4 (2) of planning data for the selected sites consists of Lat/long,
28 Height of Antenna, Azimuth, BCCH frequency, Cell ID etc
in TABLE I.
(iii) Rural path loss, TABLE I
PLANNING DETAILS OF 2G SITES
L (dB) = L(urban) - 4.78 (log fc)2 + 18.33 log Site name Azimuth Long Lat
Ant.
ARFCN
Height
fc + 40.94 (3)
15 37 75
KABLANA 165 76.747 28.621 37 71
For urban area divided into: 260 37 87
0 38 111
BUPANIA 120 76.824 28.636 35 73
For large cities: 190 38 71
E = 3.2 [log (11.7554hm)]2 – 4.97 fc 400 MHz 350 35 76
DABHODA
E = 8.29 [log (1.54hm)]2 – 1.1 fc 200 MHz KHURD
80 76.825 28.662 35 71
180 35 87
0 35 74
For small and medium-sized cities: GUBANA MAJRA 120 76.845 28.603 35 80
240 35 88
As like any other empirical propagation models, Hata model IV. OPTIMIZATION OF MODEL
contains three basic elements [5, 6]:
Initial offset parameter, Figure 1 shows the optimization process to optimize the Hata
Initial system design parameter, propagation model. Optimization process was obtained by
Establish slope of the model curve. using least square method. The optimized model will be
validated based on sub urban areas of Haryana. Statistical
L (dB) = 64.54 + 22.16 log (fc) – 13.82 log (hte) + [44.9 – analysis such as relative error, standard deviation and
6.55 log (hte)] log (R) –E (5) variance were used to compare between the optimized model
and other known models.
From Eq (5), the initial offset parameter is fixed at:
E0 = 64.54
From Eq (5), the initial system design parameter may be
expressed as:
Esys = 22.16 log (fc) –13.82 log (hte)
The slope of model curve is a constant as expressed from
Eq. (5) as follow:
sys = [44.9 –6.55 log((hte)]
III. METHODLOGY
The optimization process was obtained by using least square Fig. 1. Optimization flow diagram
method. The theoretical path loss is to be compared with the
measured data in the field. As the suburban area is to be
optimized the BTS should be selected in the different Algorithm for Optimization
suburban areas of Haryana. The optimization process can be For the optimization into the Hata model, initial system
done with the help of data extracted from the various design parameter Esys and slope parameter sys are
measured log files recorded in the field with the TEMS Kit. considered [8].
The accuracy of the result will be based on as many data of In this paper, a new appropriate tool is proposed in
COMO209-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO209-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
The results from the different exported log files for each 2G
BTS site in processed in excel as discussed in previous
chapter. The MS Excel for each site is processed with
Matlab program as per eq. (5) for the calculation of values a
and b in eq. (9). The value for each site for a and b is
recorded in a TABLE III as shown in
TABLE III
VALUES OF A AND B
Site name Long Lat Value of a Value of b
KHERI JAT 76.7853 28.5971 127.44 15.7066
KABLANA 76.7474 28.6214 123.211 11.4782
BUPANIA 76.8248 28.6362 123.975 12.9398
DABHODA KHURD 76.8259 28.6628 124.572 7.5206
GUBANA MAJRA 76.8459 28.6033 123.227 8.2963
SAHU 75.8327 29.4936 133.073 12.1442
DABWALI-2 74.6984 29.9561 134.098 17.1094
DABWALI RD 75.024 29.5522 130.229 14.0702
DABWALI-3 74.7102 29.964 130.8 6.3378 Fig. 2. Sahuwala BTS Pathloss Plot
OTTU 74.8883 29.5008 126.345 7.1927
(10)
V. COMPARISON OF RESULTS
COMO209-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
RELATIVE ERROR
SNO NAME OF BTS Modified Hata Empirical Hata
Model Model
1 BUPANIA 0.0532 0.1045
2 DABWALI ROAD 0.0162 0.1972
3 GUBANA MAJRA 0.0648 0.0827
4 KABLANA 0.0586 0.0723
5 JANDWALA JATTAN 0.0356 0.1157
6 HABUANA 0.0506 0.1562
7 KALANWANLI 2 0.0053 0.1473
8 OTTU 0.0416 0.0919
9 SWANTKHERA 0.0282 0.1684
10 DOHBA KHURD 0.0507 0.1416
AVERAGE 0.04048 0.12778
VARIANCE 0.000033 0.000149
STANDARD DEVIATION 0.00574 0.01222
The actual path loss results are compare with the original
Empirical Hata Model as well as Modified Hata model with
new parameters value as Eq.(10). The results are compared
Fig. 5. Sawantkhera BTS Pathloss Plot as shown in the Table IV.
REFERENCES
[1] Andrea Goldsmith “Wireless Communications”, Cambridge
University Press 2005,pp 1-2.
[2] Micheal, D. Y., "Foundation of Mobile Radio Engineering",
CRC Press Inc., 2000.
[3] M. Hata, “Empirical formula for propagation loss in land mobile
radio services,” IEEE Trans. Veh. Technol., vol. VT-29,, Aug.1980,
Fig. 6. Gubana Majra BTS Pathloss Plot pp. 317–325.
[4] John S. Seybold "Introduction to RF propagation", Willy
Puplisher,2005 pp 11-12.
[5] Micheal, D. Y., "Foundation of Mobile Radio Engineering", CRC
V. Statistical Analysis of Modified Model Press Inc., 2000.
[6] Jacques, L. and S. Michel, "Radio Wave Propagation Principles and
The modified model is verified for other sub urban sites Techniques", John Wiley & Sons Ltd., 2000.
as shown in above section. Statistical analysis such as [7] K.Ayyappan, P. Dananjayan, "Propagation Model For Highway In
Mobile Communication System", UBICC journal,2009.
relative error and standard deviation is calculated from [8] M. B. Roslee and K. F. Kwan, "Optimization of hata propagation
the measured log files both for Modified model as well prediction model in suburban area in malaysia," Progress In
as Original Empirical Okumura Hata Model .The Electromagnetics Research C, Vol. 13, 91-106, 2010.
average relative error for sites 0.04048 and standard [9] Ericsson Radio Systems AB, TEMSTM CellPlanner 3.4 User
Guide, 2001.
deviation is 0.00574 for the for Modified Hata Model [10] http://www.wifinerd.com/wifi-calculators.html.
and the average relative error for sites 0.12778 and
[11] http://reference.mapinfo.com/software/mapinfopro/english/9.5/ Map
Info Professional Supplement.pdf.
COMO209-5
1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO210-1
2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO210-2
3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
VI. REFERENCES
[1] Jia-Shen G. Hong & M.J. Lancaster, “ Microstrip Filters
for RF/ Microwave Applications” John Wiley & Sons
Inc., 2001.
[2] D.M.Pozar, “Microwave Engineering, “John Wiley,2000.
[3] Jia-Sheng Hong, Lancaster M .J, “ Recent progress in
Planar microwave filters,” IEEE Trans. Antenna
Propagat, Vol. 2, pp. 1134 – 1137, August 1998.
[4] G. Mathaei, L.Young & E.M.T. Jones, “Microwave Filter
Impedance matching networks and coupling structures,”
Artech House, Norwood, MA, 1980.
[5] IEEE Trans.vol.CT-5, January 1964.
[6] IE3D Software Release – 8, Developed by M/S Zeland
Software Inc.
[7] Tae - Yeoul Yun and Kai Chang, “Transaction on
COMO210-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 1
COMO211-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 2
In this paper, filters are optimized for high performance and an Өj = ∏ -1/2[tan-1 (2 BJ,J+1/Y0 ) + tan-1 (2 BJ-1,J/Y0)] radians
efficient and cost effective fabrication methods is proposed. (5)
Microstrip technology is used for simplicity and ease of Where the BJ,J+1 and Өj are evaluated at fo. Note that the
fabrication. The design and simulation are performed using 3D second term on the right hand side of second equation
full wave electromagnetic simulator IE3D. This filter is widely indicates the absorption of the negative electrical length of the
used today in radar, satellite and terrestrial communication J-inverters associated with the jth half wavelength resonator.
applications.
As referring to the equivalent circuit of microstrip gap, the
coupling gaps sj,j+1 of the microstrip end- coupled resonator
filter can be determined as
II.MICROSTRIP FILTER DESIGN C g j,j+1 = BJ,J+1/ωo (6)
An nth order filter will use N resonant sections of transmission
line With N+ 1 capacitive gap between them. The Where ωo = 2∏ fo is the angular frequency at the midband.
discontinuities may be approximated as Series capacitors. The
resonators are approximately λ/2 long at the resonant The physical length of resonator are given by
frequency ω0.
lj = (λg0/2∏ ) Өj - ∆ lje1 - ∆ lje2 (7)
Where each open end microstrip resonator is approximately a ∆ lje2 = (ωo C p j,j+1 /Y0)(λg0/2∏ ) (9)
half guide wavelength long at the mid band frequency f 0 of
the band pass filter [1]. The coupling from one resonator to the III.SIMULATION RESULTS
other is through the gap between the two adjacent open ends,
and hence is capacitive. The gap can be represented by the In order to verify the validity of the above expressions in
inverters [1]. These J-inverters tends to reflect high impedance millimeter wave regime, a simulation study was performed
level to the end of each of half wave length resonator, and it using IE3D. To get the exact response for our purpose,an
can be shown that this cause the resonator to exhibit a shunt optimization was performed using software.The dimensions of
type resonance . Thus, the filter under consideration operates the filters are given in the table and the simulated filter
like the shunt resonator type of filter whose general design responses are depicted below.
equations are:
Jj,j+1/Y0 = (∏ FBW/2)1/√ (gjgj+1) j = 1 to n-1 (2) DIMENSIONS OF THE FILTERS FOR n=3 (in mm)
COMO211-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 3
2. End Coupled l1=l3=12.56 W= s01=s34=0.03 the Butterworth band pass filter, shown in figure 2 & 4. The
Filter at l2=13.32 1.1 s12=s23=0.601 simulated response of the conventional geometry in figure 3 &
6GHz 5, the attenuation decreases at 10 GHz frequency shows in
figure. for the band edge frequency of 9.66 GHz and 11.34
GHz, we getting a sharp cutoff at 10.5 GHz & maximally flat
response without any ripples in the pass band with non
attenuated frequency in the desired band. Considerable
improvements in the frequency response can be seen.
V. CONCLUSION
Fig.2. Layout of end coupled BPF on a substrate with a The filters are the very essential part of the microwave system,
not only in microwave but they are also very important in
relative dielectric constant of 4.4 and a thickness of 1.6mm
communication field. Any communication system cannot be
designed without filters. The simulated end coupled band pass
filters achieved an insertion loss of less than 1.0dB and
10.5Ghz filters is designed on 1.6 mm thick substrate for εr =
4.4 with loss tangent 0.0024. And 6Ghz filters is designed on
1.27 mm thick substrate for εr = 4.4 with loss tangent 0.0024.
As frequency of operation increases return loss decreases.
Considerable improvements in the frequency response can be
seen.
ACKNOWLEDGEMENT
The authors would like to thank authorities of Krishna
Institute of Engineering & Technology for all the support
Fig.3.Full-wave EM simulated frequency response of the filter
provided.
for εr = 4.4 and h =1.6.
VI.REFERENCES
[1]. Jia-Shen G. Hong & M.J. Lancaster, “Microstrip Filters
for RF/ Microwave Applications” John Wiley &
SonsInc.,2001.
Fig.4. Layout of end coupled BPF on a substrate with a
relative dielectric constant of 4.4 and a thickness of 1.27mm [2]. D.M.Pozar, “Microwave Engineering,”John Wiley,2000.
[3]. Jia-Sheng Hong; Lancaster M.J, “Recent progress in
planar microwave filters,” IEEE Trans. Antennas Propagat.,
Vol. 2, pp. 1134 – 1137, August 1998.
[4].G. Mathaei, L.Young & E.M.T. Jones, “Microwave Filter
impedance matching networks and coupling structures,”
Artech House, Norwood, MA, 1980.
[5]. IEEE Trans.vol.CT-5, January 1964.
[6].IE3D Software Release – 8, Developed by M/S Zeland
Software Inc
.
[7].Tae-Yeoul Yun; and Kai Chang, “Transaction on
Microwave Theory and Techniques,” IEEE Trans, Vol 49,
Fig.5.Full-wave EM simulated frequency response of the No.3, March 2001.
filter for εr = 4.4 and h =1.27.
[8].M.Makimoto; and S.Yamoshita, “Microwave Resonators
IV. RESULTS & ANALYSIS and Filters, “IEEE Trans.Wireless communication Vol.2,
August 1986.
The conventional geometry of the chebyshev band pass filter,
shown in figure 1. the simulated response of the conventional [9].H.Ozalki and J.Ishii, “Synthesis of transmission -line
geometry in figure 2 & 4. The attenuation decreases at 10 GHz networks and the Design of UHF filters,” IRE Trans. On
frequency shows in figure 3 & 5 for the band edge frequency circuit theory, vol. CT-2, pp.325- 336; December 1955.
of 9.66 GHz and 11.34 GHz(at 6 Ghz the band edge frequency
of 5.85 GHz and 6.15 GHz). The conventional geometry of
COMO211-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 4
COMO211-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Abstract: Recently, payment card issuers in the credit card and the reader. These same
the United States have begun mass deployment features, however, are also the basis for our
concern about security and privacy
of radio-frequency enabled payment cards. We
vulnerabilities. Traditional credit cards require
examine a few examples of this new class of that an entity have visual access or direct
payment card and observe that while the card physical contact in order to obtain information
issuers have implemented some new security from the card such as the cardholder's name and
features, all of them admit practical attacks to a the credit card number. By contrast RFID credit
greater or lesser degree. This paper serves as the cards make these and other sensitive pieces of
data available using a small radio transponder
first of two installments describing the threats
that is energized and interrogated by a reader. In
and vulnerabilities of RFID-based credit cards the remainder of this technical report we will
as demonstrated by our laboratory experiments examine what data are conveyed, what kind of
and also proposed a model to make card more equipment is required to receive such a
secure. transmission, and whether such transmissions
can be initiated by adversarial (not authorized
Keywords: RFID, credit cards, contactless, by credit card companies) readers.
vulnerabilities, proposed model
2. Summary of Findings: We examined
1. About RFID Credit Cards: An increasing representative RFID enabled credit cards from
number of credit cards now contain a tiny the major payment associations. We engaged
wireless computer known as an RFID chip each of the cards in various transactions with
(Radio Frequency Identifier) or a contactless several different kinds of RFID readers,
smart card chip. There are reportedly over 20 including specialized point-of-sale equipment
million RFID credit cards already deployed in deployed by major retailers. We determined that
the U.S., and this number is increasing rapidly. cheap off-the-shelf hardware and software are
According to Visa ``This has been the fastest sufficient for an adversary with only modest
acceptance of new payment technology in the technical skills to obtain critically sensitive data
history of the industry'. RFID credit cards are from the RF interface of the cards. While it is
growing in popularity because they permit possible that some existing RFID credit cards
contactless payment transactions which are fast, have mechanisms for protection of sensitive
easy, can be more reliable than magstripe information, cards from most issuers reveal all
transactions, and require only physical of the following information totally unprotected
proximity (rather than physical contact) between by any cryptographic security mechanism:
COMO212-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Cards from one issuer revealed this same 3.1 Attacks and scenarios: The RFID
information with the exception that the credit security and privacy community defines several
card number used by the RF interface is a conventional adversaries by their methods and
different number from that encoded onto the goals. Our informal analysis considers four
magnetic stripe. Off-the-shelf hardware requires basic attacks to privacy and security.
reasonably close proximity (on the order of Clandestine bearer tracking violates a tag
10cm) in order to read these data from an RFID bearer’s privacy by exceeding expected
credit card. This is sufficient to read a credit interactions. Clandestine scanning allows an
card through clothing or a wallet, but the reader adversary to energize and read tags from either
must still be close to the targeted card. The close proximity or a distance. The replay and
maximum potential range available seems to be relay attacks allow a more powerful man-in-
hotly debated in press accounts of the situation. the-middle adversary to impersonate a card
Experiments conducted by Royal Dutch Shell of holder. Finally, a cross-contamination attack
Canada and reported in indicate a read range of allows an adversary to use information from RF
26 inches, while retailer 7-Eleven insists that the transmissions to attack non-RF media such as
range is only 2 inches. This longer range is magstripe. This study does consider more
supported by the academic literature: detailed advanced attacks such as side-channel analysis
instructions on how to build and operate such a and physical probing.
specialized reader are already available to the
public on the World Wide Web. Clandestine bearer tracking: In this scenario,
Additionally the range at which it is possible to a legitimate merchant exceeds the expected use
detect the presence of a tag or eavesdrop on a of their RFID credit card readers. For example,
transaction may be much longer than the range a merchant A may want to know whether a
required for an active skimming attac. This fact credit card C has been used with any other
is apparently behind the report from the vendor since the last transaction between A and
National Institute of Standards and Technology C. Avoine describes examples of how such
that the protocol used with these kinds of RFID attacks on the cardholder’s privacy could be
chips can be read from up to 30 feet away using implemented. Some of the cards we examined
specialized reading hardware. admit this sort of attack by means of a
transaction counter that could be co-opted.
3. Threat Model: RFID-enabled credit cards
include several noteworthy features that open Clandestine scanning: In this attack an
vulnerabilities not found on traditional unauthorized and potentially clandestine reader
magstripe cards. For instance, RFID cards can reads tags from either close proximity or from a
receive external power, communicate distance. One such scanning attack we call the
wirelessly, include persistent state that is “Johnny Carson” attack. In a famous series of
comedic skits, Johnny Carson’s character
COMO212-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
representative samples of many different RFID of PICC transmissions revealed that all PICC
credit cards from each of the three largest communications also conform to the ISO-
payment associations in the U.S. Our 14443-B protocol. Our hardware and software
conclusions are based on observations of this are able to capture and demodulate any data
sample set of approximately 20 credit cards transmitted between RFID credit card readers
obtained by five separate researchers from and cards that are within a certain distance of
several different issuing banks and three major the eaves- dropping antenna. Since the focus of
payment associations. All of these cards were this work is not on extending read ranges, we
issued within the past year. Given the size and did not try to achieve great range. But we did
diversity of our sample set we believe that our experimentally demonstrate that eavesdropping
results reflect the current state of deployed with our setup is effective through materials
RFID credit cards, however we wish to such as cloth, lending credence to the threat of
emphasize that card issuers continue to clandestine eavesdropping, perhaps through
innovate, and as time goes by we hope and clothing. Examination of data obtained through
expect that cards will change and additional these means immediately demonstrated the
security features will be added. Therefore we do efficacy of the simple eavesdropping attack,
not claim that our findings are exhaustive, in since the full cardholder name and card
particular there may already exist cards that use expiration date were present in clear text in all
security mechanisms beyond any we have transactions. Other data such as credit card
observed. number are discussed in Section 5.
RFID credit cards, we built a device capable of the size and shape of its namesake stick of
impersonating our commercial RFID credit card chewing gum, and incorporates an ARM
readers. We required only the ability to send PXA255 microprocessor . Four General Purpose
arbitrary bytes according to ISO 14443 layers 2 Input/Output pins (GPIOs) of the
and 3, and we discovered that the Texas microprocessor are used to control simple radio
Instruments s4100 Multi-Frequency RFID circuitry of our own design. The small size and
reader possessed all of the hardware capabilities low power requirements of this single board
that we required. This hardware together with computer contribute the feasibility of
applications of our own design allows us to clandestine use of a PICCAL. Our analog front
rapidly challenge cards at a rate far exceeding end is essentially a simple AM radio consisting
that observed on any commercial hardware. In of three integrated circuits, and a few capacitors
addition, this same hardware combined with and resistors. The integrated circuits are: a
custom amplifiers provides the basis for some comparator used to demodulate AM commands
noteworthy read range extension experiments. from the PCD, a counter/divider to divide the
Using libraries of our own design, we wrote a input carrier into subcarrier and baud rate
program which simply sends the exact bytes that clocks, and an XOR to allow the microprocessor
we captured from the commercial readers in our to accomplish phase shifting for the 14443-B
eavesdropping experiments. Eavesdropping on layers 2 and 3 specified binary phase shift
transactions between our credit card reader keying.
emulator and real RFID credit cards
demonstrated that all of the RFID credit cards
we tested responded to our emulator exactly as
they respond to a commercial RFID credit card
reader. This strongly suggests that cards operate
in a “promiscuous mode” interacting with any
reader, with no cryptographic or other secure
mechanisms in use to authenticate an authorized Table 1. A summary of the four types of cards
RFID reader to a credit card. and susceptibility to various attacks.
4.3 Replay experiments: An RFID credit card We programmed our PICCAL to expect the
belongs to a class of RFID devices known in the RFID credit card reader commands that we
ISO standard as a “proximity integrated circuit captured using our eavesdropping setup
card” (PICC). Since the primary difference described in Section 4.1, and to transmit replies
between our device and a traditional PICC is captured from real RFID credit cards during a
that ours uses actively powered logic circuitry skimming attack performed with the reader
as opposed to the passively powered (antenna emulator described in Section 4.2. The output
powered) RFID credit card, we have named our from our commercial RFID credit card readers
device a PICCAL (Proximity Integrated Circuit is identical in the case where the reader is
Card with Active Logic). Our PICCAL is a presented with a real versus PICCAL emulated
microprocessor controlled device capable of credit card. Since the data thus output is the
sending arbitrary bytes over the ISO 14443-B same data we would expect to be transmitted
transport layer (layer 2). For ease of prototyping over the charge processing network, we cannot
and flexibility of experimentation we chose the think of a scenario in which the charge
gumstix single board computer as the controller processing network could distinguish a real card
for our device. This computer is approximately from a PICCAL unless additional elements are
COMO212-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
present that we have not been able to observe in Card A protocol: When presented (RF
the laboratory. But as noted above, many pieces transaction) with any sample of a card from
of data go into an overall transaction approval issuer A, our reader outputs serial data identical
decision including sophisticated risk-based to the data contained on the magstripe of the
fraud detection mechanisms on the backend. For same credit card. This finding was confirmed by
this reason, a valuable future research direction comparison with output obtained with
would include field tests in which PICCAL presentation of the magstripe rather than RF.
initiated transactions are tested with complete When presented with the same card, the output
purchases from real merchants. is always the same: there is no evidence of a
counter, one-time-password, or any other
5 Analysis and Results: To protect the mechanism for prevention of replay attacks.
identity of our cards, we label the cards A, B Figure 2 shows a sample of this serial output,
and C based on semantic equivalence classes which includes all the usual components of an
determined by observing behavior between ISO 7813 magstripe. The first line represents
cards and readers. Table 1 summarizes the Track 1. The start sentinel B is followed by the
vulnerabilities of four classes of cards. primary account number. Following the field-
separator character, the cardholder name
5.1 RFID credit card protocols: In a appears, followed by another field-separator and
traditional card-present transaction, data is read an “additional data” field. This field includes
from the magnetic stripe of a credit card by a not only the card expiration date (in this case
POS terminal. The format of this data is 06/2009), but also a long string of digits. The
specified by ISO 7813. meaning of these additional digits is not clear,
In this section we shall explore some of the but since this field is static for card type A, it
RFID credit card protocols that are in current cannot be used to prevent a replay or cross-
deployment. We shall examine some of the contamination attack. The second line represents
conclusions that can be reached through standard Track 2 data, which is largely similar
examination of the ISO 7813 data output by the to the Track 1 data. Track 2 does not contain the
serial port of RFID credit card readers when cardholder name, and contains less room for
presented with different types of credit cards. proprietary information.
Where pertinent we shall consider in correlation
with this serial output the raw RF data from the {Bxxxxxx6531xxxxxx^DOE/JANE^09061010
same transactions as captured by our 00000000000000000000000000858000000}-T1
eavesdropping apparatus. In keeping with a {xxxxxx6531xxxxxx=09061010000085800000}
philosophy of ethical attacks research, we have -T2
redacted several pieces of information from the
following subsections in part due to a desire to Fig. 2. Serial output from a commercial reader
prevent criminal misuse of our findings. after an RF transaction with a card from issuer
Cardholder name and card number have been A
concealed. Additionally we have obscured the
number of digits in the card number in order to Card B protocol: When presented with cards
obscure which observations correlate with the from issuer B, our commercial readers output
products of specific payment associations and data similar to that of the card A experiments,
issuing banks. with a few important differences. In the sample
card B output shown in Figure 3 we note the
presence of a counter, determined to be such
COMO212-6
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
because of monotonic incrementation with that the counter value in Track 2 is followed by
successive transactions. Additionally we a 0 instead of a 1, perhaps an indication to the
observe three digits which change with each back end processing network of a different
transaction in no pattern that we have identified. algorithm.
Because of the relatively high entropy of these
three digits, we consider it likely that they are {Bxxxxxx2892xxxxxx^DOE/JANE
the output of some cryptographic algorithm 017^1001101010691958}-T1
which takes the transaction counter as an input. {xxxxxx2892xxxxxx=100110101069195801700
If this is the case, then the algorithm must also }-T2
take a card-specific value like a cryptographic
key as an input since we observe that different {Bxxxxxx2892xxxxxx^DOE/JANE
cards with the same counter value produce 018^1001101040146036}-T1
different codes. We speculate that these data {xxxxxx2892xxxxxx=100110104014603601800
may serve as a stand-in for the traditional CVC. }-T2
Fig. 4. Sample output from an issuer C card
{Bxxxxxx1079xxxxxx^DOE/JANE^09011011 differs from output of an issuer B. Transaction
00000000000100000000000}—T1 codes are shown in bold italic font, transaction
{xxxxxx1079xxxxxx=09011011000001600221} counter is shown underlined.
—T2 Proposed Model: - We have studied all three
{Bxxxxxx1079xxxxxx^DOE/JANE^09011011 protocols and we observe that card c protocol is
00000000000100000000000}—T1 the most secure compare to other. But in some
{xxxxxx1079xxxxxx=09011011000007400231} condition card c protocol will also be fail. What
-T2 are this condition and what improvement we
have to do in this card to make it more secure?
Fig. 3. Sample of reader serial output after RF Failure condition: Because here CODE is equal
transaction with a card from issuer B. In this to sum of key and counter value. That producing
sample we see a three digit code (shown in bold the same code for all cards at same counters
italic font), and a four digit counter (shown value.
underlined).
Example- If we have three cards and all cards
Card C protocol: Card C’s protocol differs have a same transaction and have a same
from Card B’s in a few crucial details: counter value than the entire three cards will
getting the same code value. This is also a
1. Its unique transaction codes are eight digits sophisticated problem
instead of three
2. its transaction counter, now located in the
Cardholder Name field, displays only three
Digits instead of four
3. Rather than sending the embossed card
number over the air, it uses a fixed pseudonym
Shown in Figure 4 are transactions 017 and 018
from an issuer C RFID credit card. These
transactions correspond to codes 10691958 and
40146036, seen both at the end of Track 1 and
in different order at the end of Track 2. Note
COMO212-7
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
(CODE=KEY+COUNTER VALUE)
{Bxxxxxx2892xxxxxx^DOE/JANE
017^1001101010691958}—T1
{xxxxxx2892xxxxxx=10011010CODE0170}—
T2
10 Conclusion: Despite the millions of RFID-
CODE=KEY+X enabled payment cards already in circulation,
and the large investment required for their
X=CVC VALUE OF CARD + COUNTER manufacture, personalization, and distribution,
VALUE all the cards we examined are susceptible to
privacy invasion and relay attacks. Some cards
This proposed model provides the different may be skimmed once and replayed at will,
codes to all those cards, which operate at the while others pose a modest additional
same counter value. Because each cards have a synchronization burden to the attacker. After
reverse engineering the secret protocols between
RFID-enabled credit cards and readers, we were
able to build a device to mount several advanced
replay attacks in laboratory conditions. While
absolute security and privacy in a contactless-
card form factor is difficult to achieve, we hope
that next-generation RFID-enabled payment
systems will protect against the vulnerabilities
that our study identifies.
References
different CVC value.
(PROPOSED MODEL) 1. Adida, B., Bond, M., Clulow, J., Lin, A.,
Murdoch, S., Anderson, R., Rivest, R.: Phish
and chips: Traditional and new recipes for
8 Comparative Analysis :
attacking EMV. Technical report, University of
Cambridge Computer Laboratory (2006)
Factors Card A Card B Card C Proposed
Protocol Protocol Protocol Protocol http://www.cl.cam.ac.uk/˜mkb23/research/Phish
Security Less Moderate Moderate Most and-Chips.pdf.
Secure
Code Less Moderate Moderate Most
Reliability Reliable
COMO212-8
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO212-9
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
achieving a telemetry distance of 1.5 meters transmission laser to vaporize the sample.
under operating temperatures up to 250oC. The 7mW of power available should be
sufficient.
II. SENSOR TECHNOLOGY
The initial design decisions of sensor E. Magnetometer
technology to use a single type of node, it is Magnetometers typically require a
required that every node carry a complete magnetically active element which moves
sensor package. For a probe with the according to the prevalent magnetic field,
envisaged dimensions, the overall sensor the movement being sensed using any of the
package will need to be designed in an established MEMS pick-off techniques. The
integrated way, allowing for reuse of as magnetically active element can be provided
many subsystems as possible. As each using either a permanent magnet using
sensor type is discussed below: electromagnetism, which consumes current.
that the spectrum analyzer can receive an power high-temperature MEMS sensing and
incoming signal with an SNR of at least 10 data telemetry applications. The prototype
dB over telemetry distances of 1.5 m. Fig9 wireless sensing and communication module
shows the corresponding received power achieves high-temperature operations up to
spectrum at 1.5-meter telemetry distance 250oC over a telemetry distance of 1.5
from the prototype oscillator operating at meters with a transmitter power
250oC. An extended communication range is consumption of 60 μW. Together with
expected through using a more sensitive MEMS actuation technologies, the sensing
receiver. technologies allow for new adaptive
A high-temperature frequency variation over systems, providing more efficient
time has been observed in the current aerodynamics and hydrodynamics. MEMS
prototype. The prototype exhibits an initial sensors have already proven to be a potent
frequency decline of approximately 150 kHz technology in terms of size and cost
over 30 minutes, then a random frequency reduction of everyday sensors. When
variation of 20 kHz, thus limiting the system integrated with processing capability they
resolution. can handle a range of further applications, in
particular being more readily deployed and
adapted to a specific application problem.
REFERENCES
COMO213-6
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Abstract- Mobile ad hoc networks (MANETs) are There are various ad hoc routing protocols which
collections of mobile nodes that can dynamically form use route hop count as the routing metric. In
temporary networks on the fly without the need for shortest path routing, nodes on shortest path get
pre-existing network infrastructure or centralized more heavily loaded than others because they are
administration. Due to unorganized connectivity and chosen frequently as routing path. Having a
dynamic topology, routing in MANET becomes a heavy load, constraints like lower capacity of
challenging task. Moreover, constraints like lower
capacity of wireless links, error-prone wireless
wireless links, error-prone wireless channels,
channels, limited battery capacity of each mobile node limited battery capacity of each mobile node etc.,
etc., degrade the performance of MANETs routing degrade the performance of MANETs routing
protocols. Heavily-loaded nodes may cause congestion protocols. Heavily-loaded nodes may cause
and large delays or even deplete their energy quickly. congestion and large delays or even deplete their
In this article, we discuss the various load metrics and energy quickly. Therefore, routing protocols that
summarizes the principles behind several existing load can evenly distribute the traffic among mobile
balanced routing protocols. nodes and hence can improve the performance of
MANETs are needed.
I. INTRODUCTION This article is organized as follows. The
A mobile ad hoc network (MANET) is a kind of next section categorizes various types of load
wireless ad hoc network, and is a self- metrics and load balanced ad hoc routing
configuring network of mobile routers (and protocols. Afterthat, we will compare various
associated hosts) connected by wireless links – load metrics. We then discuss load balancing
the union of which form an arbitrary topology. techniques used by existing load balanced
The routers are free to move randomly and routing protocols. This is followed by a
organize themselves arbitrarily; thus, the comparison of these various protocols in the
network's wireless topology may change rapidly following section. The final section concludes
and unpredictably. They have features of fast the article.
deployment and able to communicate without
link breakage while on the move. They can II. CATEGORIZATION
easily communicate anywhere at any time. Over Over the years, various load balanced ad hoc
the years, numerous routing protocols have been routing protocols have been proposed. Most of
developed for ad hoc mobile networks. Routing the approaches are on-demand-based protocols;
protocols in MANETs are classified into three that is, they follow the load balancing strategies
categories: table driven proactive, on-demand with route discovery [2]. A route with the least
reactive source initiated and hybrid routing [3]. load among multiple possible routes from source
Table Driven Proactive routing protocols to destination is generally chosen. As shown in
maintain consistent, up-to-date routing Fig. 1, these routing protocols can generally be
information in each node by propagating route categorized into three types (based on their load
updates throughout the whole network. balancing techniques):
Although a route to every other node is always 1 Delay-based: Where load balancing is
available, such protocols incur signaling traffic achieved by attempting to avoid nodes
and power consumption overhead. On the other with high link delay. Examples of delay
hand, on-demand reactive routing protocols do based load balanced routing protocols is
not maintain routing information at every node. Load-Aware on-Demand Routing
They create routes only when desired by the (LAOR) [6].
source. It has been shown in [4] and stated in [5 2 Traffic-based: Where load balancing is
that reactive routing protocols can perform better achieved by evenly distributing traffic
than proactive protocols in MANETs. load among network nodes. An example
protocol using this approach are
Associativity Based Routing (ABR) [7],
COMO214-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO214-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
approach reduces delay due to the efforts made upstream node propagates an error message to
in selecting the best route. the destination. The destination node then selects
LAOR implements congestion monitoring an alternative least loaded route stored in its
during the route discovery process. A node is cache to send data. Using this technique, traffic
considered congested if the total delay of a node over the network tends to be distributed in the
is greater than the upstream node or the queue long term.
length is more than 80 percent of its buffer size.
When a node becomes congested, it discards the B. HYBRID BASED PROTOCOLS
RREQ message. This approach reduces the CSLAR is an extension of Dynamic Source
routing overhead and distributes the load evenly Routing (DSR), which also uses route discovery
in the network. and route maintenance.
CSLAR utilizes the contention information
B. TRAFFIC BASED PROTOCOLS obtained from the 802.11 distributed
ABR uses the property of “associativity” to coordination function (DCF) to take into account
decide on which route to choose. In this channel contention from neighboring nodes.
algorithm, routes are discovered by broadcasting Load balancing in CSLAR is also performed
a broadcast query request packet; with the during route discovery. When a node has packets
assistance of these packets, the destination to send, a route request packet is flooded
becomes aware of all possible routes between throughout the network. Route request is also
itself and the source. Based on these available forwarded by intermediate nodes after they have
routes, a path is selected using the associativity appended their load values to the request
property of these routes. message. However, intermediate nodes are not
The ABR algorithm maintains a “degree of allowed to send route replies back to the source
associativity” by using a mechanism called even if they have routes to the destination in their
associativity ticks. According to this, each node route caches. The destination node then decides
in the network maintains a tick value for each of to reply to route requests based on the
its neighbors. Every periodic link layer HELLO comprehensive route load value. The destination
message increases the tick value by one each node replies to the first request or to the request
time it is received from a neighbor. Once the tick that has a smaller route load than previous
value reaches a specified threshold value, it requests. Once the source node receives the
means that the route is stable. If the neighbor destination node’s reply, it then utilizes the
goes out of the range, then the tick value is reset selected route to transmit data. If one or more
to zero. Hence a tick level above the threshold links in the selected route is broken, the source
value is an indicator of a rather stable association node is notified by a route error packet, resulting
between these two nodes. Once a destination has in the source updating its route cache and
received the broadcast query packets, it has to reinitiating a new route discovery to discover a
decide which path to select by checking the tick- new least loaded route.
associativity of the nodes. The route with the ZRP is a protocol used under hybrid category for
highest degree of associativity is selected since it ad hoc mobile routing protocols. It incorporates
is considered the most stable of the available the merits of on-demand and proactive routing
routes. protocols. ZRP is similar to a cluster with the
LBAR’s load metric is similar to ABR, which is exception that every node acts as a cluster head
based on active path activity. LBAR has two and a member of other clusters. The routing zone
phases: route discovery and route maintenance. comprises a few mobile ad hoc nodes within one,
The route discovery process is initiated two or more hops away where the central node is
whenever a source node needs to communicate formed. Since ZRP uses both reactive and
with another node for which it does not have a proactive schemes, it exhibits better
route. It broadcasts the setup message to its performance. However, since hierarchical
neighbors. A node receives a setup message then routing is used, the path to a destination may be
forwards it after updating the nodal activity and suboptimal. Since each node has higher level
traffic interference information. Such topological information, memory requirement is
information is then used to calculate the route greater.
load [2]. The destination node then selects a LARA is another hybrid load balanced routing
route with the least load and sends an ACK protocol. LARA requires each node to maintain a
message back to the source node. During route record of the latest traffic queue estimation of its
maintenance, when a route is broken, the neighbors. The traffic queue is defined as the
COMO214-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
average value of the interface queue length receives the first route request. Comparing route
measured over a period of time. Traffic density, discovery delay, on-demand load balanced ad
on the other hand, refers to the sum of traffic hoc routing protocols will incur higher delay
queues at a node plus the traffic queues of all the since nodes need additional waiting time to
node’s neighbors.. LARA selects the best route discover possible routes to destinations.
based on checking the traffic type (i.e., TCP or To find the least loaded path among multiple
non-TCP traffic). For TCP traffic, route selection possible routes, ABR, LBAR, and LARA
is based first on route hop count followed by protocols allow the destination node to wait for
traffic cost. For non-TCP traffic, the route with some time (Δt) before replying with the best
the minimum traffic cost is selected. route to the source node. Meanwhile, TSA
allows the source node to wait for a time interval
III. COMPARISONS OF LOAD BALANCED ( Δt) to select the best route to the destination
AD HOC ROUTING PROTOCOLS node. Therefore, the route discovery delay is: 2T
There is a comparison of several load balanced + Δt, where T is delay incurred for the
routing protocols. The key differences among the destination node to receive the first route request
routing protocols lie in the method of load message. LAOR and CSLAR protocols,
estimation and the decision on route selection. however, incur a shorter delay since the
destination node will reply to the first route
A. LEAST LOADED ROUTE DECISION request received (Table 1), so the delay incurred
Protocols such as ABR, LBAR, and LARA use is 2T.
destination-based decision load balancing. Once
the destination node receives a route request, it D. ROUTING OVERHEAD
waits for some time to select the route with the LBAR, LARA, and TSA use control messages to
least loaded path. When the source node receives exchange load information among neighbors,
the reply message, it proceeds to send data. TSA, which is then used to calculate the load metric.
however, uses source-based decision load On the contrary, LAOR and ABR allow nodes to
balancing. In TSA multiple reply messages are compute load information (by counting the
sent from destination node to source node. After number of active paths) without incurring
waiting for a predefined interval, the source node additional overhead.
selects the least loaded path.
E. COMMUNICATION COMPLEXITY
B. ROUTE REQUEST FLOODING METHOD Communication complexity refers to the number
Most of the load balanced routing protocols are of messages needed to complete a protocol
on-demand-based protocols in which a source operation. During the initiation phase, ABR,
node discovers routes by first flooding a route LBAR, and LARA require O(N + y), where N is
request message. In the duplicate route request at the number of nodes in the network and y is the
destination node (DQD) approach, the total number of nodes forming the directed path
destination node is allowed to receive duplicate the REPLY packet transits. In TSA and CSLAR,
route requests. The destination node then decides to obtain a route from source to destination, the
to reply to all/ some/one route request. Currently, destination node must reply to all route requests.
only the LAOR protocol uses this approach for The communication complexity of TSA and
route discovery. Meanwhile, protocols such as CSLAR is similar: O(2N). The complexity of
CSLAR, TSA, and LAOR use the duplicate LAOR is O(N2) due to the allowance eof
route reply at source node (DPS) approach, intermediate nodes to receive duplicate route
where the source node is allowed to receive request messages.
duplicate route reply messages from the When mobility causes route failures (i.e., post-
destination node. failure), the communication complexities of
ABR and LBAR are the same, O(x + y), where x
C. ROUTE DISCOVERY DELAY is number of nodes affected by a topological
The delay incurred while waiting for multiple change. Meanwhile, the communication
route request or reply messages is an important complexities of LARA, CSLAR, TSA, and
factor to be considered in the design of ad hoc LOAR are O(N + y), O(2N), O(2N), and O(N2),
routing protocols. To avoid the issue of large or respectively, since these routing protocols do not
least waiting time, both CSLAR and LAOR employ localized route discovery. When
protocols allow the destination node to send a mobility causes route failures (i.e., post
reply whenever it failure), the time complexity of ABR and
COMO214-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
LBAR is O(l + z), where l is the diameter of network, and z is the diameter of the directed
the affected network segment. Since these path the REPLY packet transits. However, for
routing protocols use localized route CSLAR and LAOR, the complexity is O(2d)
discovery, time complexity therefore
IV. CONCLUSION
depends on the diameter of the affected In this paper we discuss the various types of
network segment. Meanwhile, CSLAR, load metrics and the principles behind some
TSA, LARA, and LAOR have the same time existing load balanced ad hoc routing protocols:
complexity as in route initiation since they ABR, LBAR, LARA, CSLAR, TSA, and LAOR.
reinitiate a new route discovery when route While offering better representation of actual
failures occur. Thus, the time complexities load, LARA and CSLAR incur higher
of CSLAR, TSA, LARA, and LAOR are complexity in capturing load information.
O(2d), O(d + z), O(d + z), and O(2d), Comparing the operations of routing protocols,
respectively. only LBAR and ABR perform load balancing
during route maintenance. Other protocols
F. TIME COMPLEXITY balance load only during route discovery.
Time complexity refers to the number of steps
needed to perform a protocol operation. During
route initiation, ABR, LBAR, TSA, and LARA
incur O(d + z), where d is the diameter of the
COMO214-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO214-6
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26 27 2011
COMO215-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26 27 2011
COMO215-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26 27 2011
COMO215-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26 27 2011
are no neighbor node that can algorithm for systems where nodes
advance further than the current have no location information, they
node and hence face routing designate two nodes as beacon
(perimeter routing) is to be used to nodes. Next nodes determine if they
route under these conditions. The are perimeter nodes from a heuristic
packet returns to Greedy forwarding based on their hop count from the
as soon as it finds that greedy beacons. Once the perimeter nodes
routing is available again. This is the are determined, O(p2) messages are
concept of Greedy Perimeter exchanged, where p is the number of
Stateless Routing (GPSR), which is perimeter nodes, and the perimeter
the most popular geographic routing nodes use an error-minimization
protocols. GOAFR( greedy other algorithm to compute their
adaptive face routing) is an improved coordinates. Finally, the perimeter
geographic routing that combines nodes are projected onto an
greedy forwarding with face routing imaginary circle and nodes
and achieves worst-case optimality determine their virtual coordinates
of the path length by using limited using a relaxation algorithm that
elliptic regions for face routing and works by averaging the coordinates
recursively increasing the ellipse size of neighboring nodes.
until reaching a close-to-optimal However the NoGeo system
path. assumes the system is static once
the perimeter nodes are determined.
3. Geographical Routing for Such a topology is bad for
Localization: geographic routing since there is a
For generating virtual coordinates in mismatch between the geometric
the case of geographical routing, and routing distance.
Leong et al. had earlier proposed the The novel approach to
GSpring[5] family of coordinate Localization of Wireless Sensor
assignment algorithms for adhoc nodes would be to make use of
wireless networks. Starting from a Geographical routing where virtual
set of initial coordinates derived from coordinates are identified for efficient
a set of elected perimeter nodes, routing. Hence there is a need to
Greedy Embedded Spring incorporate a technology which can
Coordinates(GSpring) detects generate virtual coordinates that
possible dead ends and uses a produces usable coordinates quickly
modified spring relaxation algorithm and improves the routing
to incrementally adjust virtual performance of existing geographic
coordinates to increase the convexity routing algorithms. For the purpose
of voids in the virtual routing of generating virtual coordinates the
topology. This reduces the possibility GSpring method is a good option.
that packets will end up in dead ends The primary focus of the work will be
during greedy forwarding. Rao et al. to develop an algorithm which can
had proposed the NoGeo family of assign greedy virtual coordinates to
coordinate assignment algorithms for the sensor nodes.
adhoc wireless networks[13].In the
most general version of their
COMO215-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26 27 2011
COMO215-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26 27 2011
COMO215-6
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO216-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO216-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO216-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO216-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
2) Performance Metrics
We evaluate the performance of TAODV, we
simulated and compared the schemes:
AODV Fig. 5 Packet Delivery Ratio v/s Mobility
TAODV.
We evaluated these schemes as a function of b) Number of total packets transmitted per data
mobility speed of nodes and the traffic of packet delivered Figures 6 shows the routing
sessions. The number of sessions was set to 1 overhead as a function of mobility speed. In
and mobility speed of nodes was varied from 0 figure 6 data rate is set to 300Kbit/s. As shown
to 100 m/s with pause-time that is equal to 5 in figure 6 as speed increases, because of link
seconds. To evaluate the performance of our break and route reconstruction the routing
protocol, each source sent data as a rate of overhead in AODV and TAODV are increased.
l00Kbit/s to 600Kbit/s. Since TAODV transmits many RREQs and
RREPs in compare to AODV thus its routing
We evaluated the following metrics for each overhead is higher than AODV.
session:
Packet Delivery Ratio: The number of
data packets received by destinations
over the number of data packets sent by
the source.
Routing Overhead. The number of all
packets (data and control packets)
transmitted divided by the number of
data packet delivered to the destination.
Average of End-To-End Delay: The
end-to-end delay of a packet is defined
as the time a packet takes to travel from
the source to the destination.
3) Performance Results
a) Packet Delivery Ratio
The packet delivery ratios as a function of
mobility speed is shown in figure 5. In figure 5
data rate is set to 300Kbit/s. We can observe that
Fig. 6 No. of packets transmitted per data
as speed increases because of links break the
packet delivered v/s Mobility
packet delivery ratios decrease in AODV and
TAODV protocols. Since in TAODV, all
c) Average of End-To-End Delay: Figures 7
discovered routes are used simultaneously for
shows the average end to end delay as a function
transmitting data, large amount of data packets
of mobility and data rate. In figure 7 data rate is
will receive to the destination in compare of
set to 300Kbit/s since TAODV uses all
AODV that use only one route for transmitting
discovered routes simultaneously its end-to-end
data.
delay is less than AODV.
COMO216-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
REFERENCES
[1] E. M. Royer and C.-K. Toh, “A Review of
Current Routing Protocols for Ad-Hoc Mobile
Wireless Networks,” IEEE Pers. Commun., Apr.
1999, pp. 46–55.
[2] A. H. Altalhi and G. Richard III, “Load-
Balanced Routing through Virtual Paths: Highly
Adaptive and Efficient Routing Scheme for Ad
Hoc Wireless Networks,” 23rd IPCCC, 2004.
[3] P. Johansson et al., “Scenario-based
Performance Analysis of Routing Protocols for
Mobile Ad Hoc Networks,” Proc. ACM/IEEE
MobiCom, Aug. 1999.
[4] ] Park, S., Shin, J., Baek, S., Kim,
S.C.,"AODV-Based Routing Protocol Considering
Energy and Traffic in Ad Hoc Networks." In Proc. of
International Conference on Wireless Networks.
(2003) 356-36.
[5] Cho, H.K., Kim, E.S., Kang, D.W."A load-
balancing routing considering power
conservation in wireless ad hoc networks” Proc.
of 16th International Workshop on Database and
Expert Systems Applications. (2005) 128-132.
COMO216-6
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Abstract— In this paper a novel Cross- ground plane or the perimeter of the patch.
Shaped patch antenna is presented.the In the case of using a high permittivity
operating frequency of antenna is 2.25 substrate, the dielectric constant and the
GHz.The dielectric constant is 4.2, resonant frequency of the antenna are related
thickness of antenna is 1.6 mm and loss as shown in (1).
tangent is 0.0013.Simulation result at
operating frequency 2.25 GHz of antenna
are done by the help of IE3D Zeland (1)
software (Version 12.0).The antenna is
small and is suitable for wireless portable
device.The bandwidth achieved is 24%. From the above equation, we can see that as
Index Terms ― microstrip antenna, the dielectric constant (εr) increases, the
cross shaped antenna. resonant frequency (fr) and the patch
antenna dimension decreases. However, it
should be noted that using high permittivity
I. INTRODUCTION substrate results in the excitation of greater
surface waves, which reduces the
performance of the antenna.
In recent years, microstrip printed antennas
have been of great interest to researchers In this paper the above-mentioned size
and engineers and have a wide spectrum of reduction techniques are investigated on
applications. The popularity is due to its key probe fed rectangular antennas. The results
attractive features such as low cost, small are then analyzed to determine which size
weight and ease of fabrication. The most reduction technique gives the highest
notable problem of patch antennas is that its reduction in the overall antenna size
size tends to be large at the low microwave compared to the conventional probe fed
frequencies. Hence, wireless applications antenna. A new cross-shaped shorted
require antenna size that is compliant with antenna for mobile communications is then
the space restrictions associated with the designed. The resonant frequency of the
portable devices. antenna is designed to be around 2.25 GHz
There are a few methods presented in the as shown in figure 1.
literature to reduce the size of a patch
antenna [1 - 3]. These include using high
permittivity substrates, loading the patch
with shorting pins and meandering the
COMO302-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
A (3)
B (4)
D
The equation (2),(3),(4) are used to
determine the length and width of a patch
antenna.
E F
= (2)
COMO302-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
IV. CONCLUSION
REFERENCES
[1] D.P. Chrissoulidis, P.C. Liakow, and
D.Y. Notis, “Dual polarized
microstrip patch antenna, reduced in size by
use of peripheral slits,”
Aristotle university of Thessalaniki, Greece.
[2] R.B. Waterhouse, “Small microstrip
Figure 4: simulation result for VSWR patch antenna,” Electron. Lett., vol. 31, pp.
604-604, April 1995.
[3] H. K. Kan and D. Pavlickovski, “Small
dual C–slot printed antenna,”
Electron. Lett., vol. 39,(7), pp. 593-594,
III. RESULTS April 2003.
[4]J.Lal & H.K.Kan, “cross shaped shorted
Figure 2 represents the design of Cross-
patch antenna”
shaped patch antenna with feed point
[5] C.A. Balanis, Antenna Theory: Analysis
(19,15).Notices that the central resonant
and Design, 2nd ed.,New York: Jhon Wiley,
frequency is near to 2.25 GHz and, in
1997
COMO302-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Abstract-- A U-slot loaded microstrip patch the use of various impedance matching and feeding
antenna is analyzed using circuit theory concept. techniques [2], the use of multiple resonators [3-5]
It is found that broadband behaviour of the and the use of slot antenna geometry [6-8]. Among
antenna is achieved due to the effect of U-slot these, the use of multiple resonators in stacked or
inserted into the patch. The optimized antenna coplanar antenna geometries is very promising
shows operational frequency band of 1.63 GHz method to enhance impedance bandwidth. However,
with impedance bandwidth of 34.72 %. A the stacked geometry increases the thickness of the
comparison between theoretical and simulated antenna on one hand and the coplanar geometry
results is also presented. increases its lateral size on the other hand. Therefore,
these geometries are not preferred for most of the
Key-words: broad band, microstrip antenna, patch modern wireless communication systems that require
antenna, U-slot single layer single patch broadband microstrip
antenna.
1. INTRODUCTION Huynh and Lee [1], in 1995, experimentally
Microstrip antennas have become attractive reported that the impedance bandwidth of single-
candidates in a variety of commercial applications layer single-patch microstrip antenna could exceed
such as mobile and satellite communications. 30 % if U-slot is cut into the patch. But the main
Traditionally, microstrip antennas suffer from low disadvantage of this antenna is its electrical thickness
bandwidth characteristic. But many of which causes additional inductive component
communication systems require wide bandwidth associated with the input impedance. In this
which is not provided by the conventional microstrip endeavor, therefore, U-slot loaded microstrip patch
antenna. Hence serious efforts started among the antenna with reduced thickness and wider bandwidth
scientific community to remove its inherent is presented. The theoretical investigation was carries
drawback of narrow bandwidth. There are numerous out using circuit theory concept based on modal
and well known methods to increase the band width expansion cavity model. The results so obtained are
of the antennas including: the use of the substrate compared with IE3D simulated [**] and
thickness [1], the use of low dielectric substrate [1],
COMO303-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO303-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
2ZUV ZUH
ZU (2)
ZUV 2ZUH
(a)
COMO303-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
where L p is the inductance due to co-axial probe of 4.108 GHz and 5.17 GHz. The two resonances are so
closely spaced that the antenna gives broadband
50 ohm.
characteristics. The calculated operational frequency
The return loss of the antenna can be calculated as
band ranges from 3.88 GHz to 5.51 GHz (i.e. band of
RL 20log (4) 163.6 MHz) and antenna shows bandwidth 34.72 %.
On the other hand, the simulated [12] return loss
Z in Z 0
where curve shows frequency band of 154.8 MHz with
Z in Z 0 bandwidth of 33.04 %. The theoretical result is in
good agreement with the simulated result. Similar
where Z0 is the input impedance of the coaxial feed
results were reported experimentally by Lee et al [6]
(50 Ω) and Huynh et al [8]. Very small errors appear in the
The radiation pattern of U-slot microstrip resonance frequencies and the bandwidth due to
approximation in calculations of the impedances of
patch antenna can be calculated as [11]
the patch and the U-slot.
k 0W
sin sin sin
jk 0WVe jk0 r 2
E ( ) cos(kh cos ) -2
r k 0W
sin sin
2 -4
(5)
k L Theoretical
cos 0 e sin sin cos Simulated
2 -6
where 0
Return loss (dB)
2 -8
k 0W
sin sin sin -10
jk 0WVe jk0 r 2
E ( ) cos(kh cos )
r k 0W
sin sin -12
2
(6)
k 0 Le -14
cos sin sin cos sin
2
-16
where 0 3.5 4 4.5 5 5.5 6
2 Frequency (GHz)
where V is radiating edge voltage, r is the distance of Fig.4 Variation of return loss with frequency
COMO303-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Simulated
-10
[1] D. H. Schaubert, D. M. Pozar and A. Adrian,
-15
“Effect of microstrip antenna substrate thickness
and permittivity: Comparison of theories and
-20 experiments”, IEEE Trans. Antenna Propag.,
vol. 37, pp. 677-682, Jun. 1989.
-25
[2] H. F. Pues and A. R. Van De Capelle, “An
-30 impedance technique for increasing the
-80 -60 -40 -20 0 20 40 60 80
Angle (degree) bandwidth of microstrip antennas”, IEEE Trans.
Fig. 5 Radiation pattern at frequency 4.517 GHz Antenna Propag., vol. 37, no. 11, pp. 1345-1354,
The calculations for radiation pattern for the Nov. 1989.
U-slot microstrip patch antenna at frequency 4.517 [3] G. Kumar and K. C. Gupta, “Broad band
GHz were done using equations (5-6), the data so microstrip antennas using additional resonators
obtained are shown in Fig.5. It is noted that the gap coupled to the radiating edges”, IEEE Trans.
calculated radiation pattern is also almost similar to Antenna Propag., vol. 32, no. 12, pp. 1375-1379,
the simulated radiation pattern. Antenna shows gain Dec. 1984.
of 5.37 dBi to 9.02 dBi in the operation frequency [4] J. A. Ansari and Ram Brij Ram, “ Analysis of
band. As shown in Fig. 6. compact and broadband microstrip patch
antenna,” MIcrow. Opt. Tech. Lett., vol. 50, no.
10
8
[5] Z. F. Liu, P. S. Kooi, L. W. Li, M. S. Leong and
7
T. S. Yeo, “A method for designing broad band
Gain (dBi)
6
microstrip antennas in multilayered planar
5 structures”, IEEE Trans. Antenna Propag., vol.
4 47, no. 9, pp. 1416-1420, Sept. 1999.
3 [6] K. F. Lee, K. M. Luk, K. F. Tong, S. M. Shum,
2
3.5 4 4.5 5 5.5 6
T. Huynh and R. Q. Lee, “Experimental and
Frequency (GHz)
simulation studies of the co-axially fed U-slot
Fig.6 Variation of gain with frequency rectangular patch antenna”, IEE Proc. Microw.
IV. CONCLUSIONS Antenna Propag., vol. 144, no. 5, pp. 354-358,
It is, therefore, concluded that the existence of 1997.
two resonance frequencies provides broad band [7] R. Bhalla and L. Shafai, “Resonance behavior of
characteristic to the U-slot microstrip patch antenna. single U-slot and dual U-slot antenna”, IEEE
The optimized antenna can be made to operate with
COMO303-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO303-6
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 1
COMO304-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 2
Where g1, g2, g3, g4......, gn .are the element of a ladder low pass Equal-ripple l1=l3=9.47 W=1.6 s01=s34=0.038
protype with a normalized cutoff Ωc = 1.0, and FBW is the BPF l2=10 s12=s23=0.81
frational bandwidth of bandpass filter. The Jj,j+1 are the
characteristic of J-inverters and Y0 is the characteristics
admittance of the microstrip line[3]. Butter-worth l1=l3=8.21 W=1.6 s01=s34=0.075
BPF l2=8.47 s12=s23=0.810
Assuming the capacitive gaps act as perfect, series-
capacitance discontinuities of susceptance Bj,j+1 as
COMO304-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011 3
ACKNOWLEDGEMENT
The authors would like to thank authorities of Krishna
Institute of Engineering & Technology for all the support
provided.
VI.REFERENCES
V. CONCLUSION
The filters are the very essential part of the microwave system,
not only in microwave but they are also very important in
communication field. Any communication system cannot be
designed without filters. The simulated end coupled band pass
filters achieved an insertion loss of less than 1.0dB and both
filters are designed on 0.672 mm thick substrate for εr = 3.02
COMO304-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
an antenna of another user with maximum SINR then the from which the base station selects M for communication. The
selection algorithm will not choose the antenna. Hence using received signal by the nth user at any timeslot is
this scheme not all the best antennas are selected. All these
scheduling schemes are implemented for independent and yn Hnx wn (1)
identical distributed (IID) channels.
In this paper, we propose an algorithm for antenna selection
using partial feedback and compared the throughput where y n is an K 1 vector, H n is a K M complex
performance with that of [12]. In the proposed scheme spatial Gaussian channel matrix from the transmitter to the nth user,
multiplexing and receive antenna selection are both employed. x is an M 1 vector of the transmitted signal and w n is the
Here we propose that each user finds out a set of M highest
SINR values corresponding to the M transmit antennas, and K 1 noise vector ~ CN (0,1) . Each of the elements in the
also the set of M indices of transmit and receive antennas channel matrix is identical and independent distributed (IID)
which achieve the highest SINR. The set of SINR values along complex Gaussian with zero mean and unit variance i.e.
with the set of indices of transmit and receive antennas are hn(k,m) ~CN (0,1) . hn(k,m) represent the complex gain from
feedback to the Base Station. The base station transmits M transmit antenna m to the receive antenna k of user n. The
independent or same message streams to M receive antennas, block fading channel model is assumed. We further assume
irrespective of user to whom they belong with the highest that the channel matrix is known at the receiver, but unknown
SINR. It will be shown that our schemes provides more at the transmitter. It is assumed that the transmitter has a power
throughput than [11], [12]. *
constraint P, i.e. E{Tr (xx )} P . We assume the total
The rest of the paper is organized as follows. Section II
transmit power is equal to unity. The total power is distributed
introduces the system mode. In Section III, a multiuser
equally among all the antennas. Hence the average transmit
scheduling algorithm is proposed for MIMO broadcast
power per antenna is 1⁄M. The power constraint is assumed to
channels and its achievable throughput is obtained. In Section
IV, results of different simulations are being presented. And be per frame. The received signal at the k th antenna of user n
the different results are compared with the simulation results of can be derived from (1) as
[12]. In Section V we presented the conclusion and scope of M
future work. y n (k ) hn (k , m) x(m) wn (m) (2)
The remainder of this paper is organized as follows. m 1
System model is given in Section II. In Section III, we x(m) is the desired signal for user n transmitted through
describe the CLLL algorithm in details. In Section IV, Atenna m with x (m) assumed to be independent for
proposed MIMO BC channel scheduling based CLLL is
individual transmit antennas. Then the signals
introduced. Simulation results and discussions are given in
x(m ), m m are considered as interference for the user.
Section V, and conclusions are drawn in Section VI.
Hence the SINR of y (n) is given as
2
II. SYSTEM MODEL (n)
hk (k , m)
SINR m,k (3)
The wireless MIMO system that is considered here has M M 2
transmit antennas at the base station and N number of users hk (k , m )
with each users having K number of receive antennas as shown n m m
COMO305-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
*
where n is the user selected by the transmit antenna m and
m 8
*
k is the selected receive antenna for the corresponding user.
m
7
The basic idea behind the above scheduling algorithm is
that each user calculates SINR for every combination of
Total Throughput (bps/Hz)
6
transmit and receive antenna and feedback the M maximum
SINR corresponding to M transmit antennas, index of 5
SINR. Since Antenna Selection is being performed at the Base SNR=0dB (Proposed)
SNR=5dB (Proposed)
Station the overhead due to processing at the receiver side is 2 SNR=10dB (Proposed)
COMO305-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
8 8
SNR=0dB (Limited Feedback)
SNR=5dB (Limited Feedback)
SNR=10dB (Limited Feedback) 7
7 SNR=0dB (Proposed)
SNR=5dB (Proposed)
SNR=10dB (Proposed)
6
6
Total Throughput (bps/Hz)
1 0
0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40
No. of Users, N No. of Users, N
Fig. 3. Comparison of Total throughput vs. the number of users between Fig. 4. Comparison of Total throughput vs. Number of users using the
the proposed scheduling algorithm and the limited feedback threshold Sth for the proposed method and the limited feedback method.
scheduling algorithm.
COMO305-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO305-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
ABSTRACT
VLSI technology is boon to our communication
technology.With the widespread proliferation of
telecommunication technology in recent years, the
need for small size multi band antennas have been
increased manifold. Hilbert antenna is one of the
example for the same .By the invention of monolithic
I.C fabrication technology i.e V.L.S.I this antenna can
find many application applications in VHF/UHF
communication . A lot of Hilbert antennas can be
designed on single substrate. This yields to a
reduction of size which is the basic need of wireless
technology. As we know the Universal truth of
resistance ,that it use to to take maximum area in IC
fabrication technology which is the main motto of
LCC viz. TEJAS. Hence, for LCC the selected radome
and antennas should be such that , it has lesser
weight & highly strong ,so that It can become
aerodynamically stable
COMO306-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
0.945
g/ o = 1.045-0.365 ln r + {6.3(w/h) r / (238.64+100w/h)}-
[0.148 – (8.81( r +0.95))/100 r] ln (h/ o)
It is clear that for all the cases studied a third order Hilbert slot
antenna can be designed with an approximate axial length, Lax =
0.80 g where g is the guided wavelength given in [2].
fig. 2.3Geometry of the Hilbert Curves for the first four Iterations
Assuming that this current distribution is sinusoidal as given by Iz It would be interesting to identify the Fractal properties of this
=I2 sin z. geometry.
So, V21= 0L E21 sin z. dz Then The plane filling nature is evident by comparing the first few
iteration of geometry shown in figure no. 2.3
There are many Hilbert curves for first little iteration as shown in D=log [(4n-1)/(4n-1-1)]/ log [(2n-1)/(2n-1-1)]
figure no. 2.3 Let us consider side by side dimension of Hilbert
Antenna is ‘L’ and order ‘n’. = (For large n) log (4n/(4n-1)/ log (2n)/(2n-1)
=2.
The length of each line segment ‘D’ is given by The similarity dimension of this curve approaches to an integer
value (2) because of the approximation involved when a large
D=L0+L1+L2+--------------------Ln = L/(2n-1) Fractal order is considered.
(Geometrical Progression)…(1)
2.7 MUTUAL COUPLING EFFECTS IN HILBERT
And the sum of all the line segments ANTENNAS
The major problem in Hilbert Antennas is mutual coupling
S=(22n-1) D=(2n+1) L …(2) between two such Antennas. Two identical Hilbert antennas can
be located in following configurations (as shown in figure no.2.5)
It should be noted that the resonant frequency is decreased as the
iteration order is increased. 1 Collinear
2.8 MUTUAL IMPEDANCE OF PARALLEL SIDE BY SIDE Let d be separation of antennas and L be the length of elements
ELEMENTS OF HILBERT ANTENNA (MATHEMATICAL (as shown in fig. 2.7)
ANALYSIS)
Ei(+-jy)= Ci(y)+-jSi(y)
Z21= R21+j X21=j30 0L[{exp (-j d2+z2)}/( d2+z2)+ {exp (-j d2+(L- The mutual resistance and reactance decreases with distance
z)2}/( d2+(L-z)2)]Sin z. dz between elements of Hilbert antenna. The exact expression where
the antenna length L is not restricted to an odd number of /2, the
=30[2 Ei(-j d)- Ei(-j ( d2+L2 +L)- Ei(-j ( d2+L2 -L)]
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
mutual resistance and reactance are given by BROWN and KING as Spacing s=h-L, CARTER gives the mutual resistance and reactance
[5] as
R21 = -15Cos h [-2Ci2 h +Ci2 (h -L) + Ci2 (h +L) -ln {(h2-L2)/ h
2
R21=30[2(2+Cos L) Ci d-4Cos2 ( L/2){Ci ( 4d2 +L2 - L)/2+ }]+15Sin( h)[2Si2 h -Si2 (h -L) - Si2 (h +L)] .
Ci ( 4d2 +L2 + L)/2}+ Cos L {Ci ( d2 +L2 - L) + Ci ( d2 +L2 + …(11)
L)}+Sin L {Si ( d2 +L2 + L)- Si ( d2 +L2 -L)-2 Si ( 4d2 +L2 + L)/2+
X21=15Sin h [2Ci2 h -Ci2 (h -L) - Ci2 (h +L) -ln {(h2-L2)/ h 2}]-
2Si ( 4d2 +L2 - L)/2}]/Sin2 ( L)/2 . (9)
15Cos( h)[2Si2 h -Si2 (h -L) - Si2 (h +L)] .
And …(12)
X21=30[-2(2+Cos L) Si d-4Cos2 ( L/2){Si ( 4d2 +L2 -
L)/2+Si ( 4d +L + L)/2}+ Cos L {Si ( d +L - L) + Si ( d2 +L2 +
2 2 2 2 Curves for R21 & X21 of parallel collinear /2 antennas and L= /2
L)}+Sin L {Ci ( d2 +L2 + L)- Ci ( d2 +L2 -L)-2 Ci ( 4d2 +L2 + are presented as function of spacing s. It has been proved that, the
L)/2+ 2Ci ( 4d2 +L2 - L)/2}]/Sin2 ( L)/2 . mutual resistance and reactance decreases with distance between
(10) elements of Hilbert antenna.
Where
C’= [{ d2+(h+L)2}-(h-L)]
It has been proved that, the mutual resistance and reactance
decreases with distance between elements of Hilbert antenna. Vr=Voltage induced at terminals of receiving antenna,
The stacked or Echelon arrangement is most general solution of which Rr=Radiation resistance of receiving antenna
the side-by-side position and the collinear position are the special
The power transmitted is given by
cases and it is highly dependent on orientation.
Pt=It2Rt (W) .
2.11 MUTUAL IMPEDANCE OF ELEMENTS OF HILBERT
ANTENNA IN INCLINED FORM ……………… (18)
Referring to the shown fig.2.9 Consider the length of the element is L
and L’ separated by a distance R with orientation angles and ’ as Where
indicated. Let us assume L <<1, L ’<<1 and r >>1. The mutual
It=Current at terminals of transmitting antenna
impedance Z21 is given by
Pr/Pt=ArAt/r2 2
……………….(20)
Where
(2) Orientation
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Figure 3
X=spacings
Y=Ohms
Graph-1
X=spacings
Y=Ohms
GRAPH-2
MUTUAL REACTANCE OF TWO PARALLEL SIDE-BY-SIDE ELEMENTS OF HILBERT
ANTENNAE
COMO306-13
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
X=spacing=Ohms
Graph 3
MUTUAL RESISTANCE OF TWO PARALLEL CO-LINEAR ELEMENTS OF
HILBERT ANTENNAE
COMO306-14
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
GRAPH-4
COMO306-15
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
1
vicky.glbitm@gmail.com, 2 amitsehgal09@gmail.com, 3 rajkecd@gmail.com
Abstract- Frequency flat, fast Rayleigh fading may be one becomes motivated to search instead for a
considered the most critical disturbance in a wireless solution that is simple in form and also simple to
communication system. In its most general form, it is modeled
as a multiplicative time continuous random (zero mean
evaluate. Another cause for this motivation is that
complex Gaussian) distortion of the transmitted signal. In the methods used to derive these alternative simple
order to obtain a reliable wireless link for time continuous forms must also be applicable to situations where
Rayleigh fading channel, each part of the system must be the closed-form solutions are not easily obtainable.
carefully designed. The special properties of the channel are There are several performance parameters of a
used as starting points in the design. Over the last 10 years a
tremendous amount of work has been presented about
digital communication system like SNR, amount of
communication over the Rayleigh fading channel. It seems fading, average outage duration, outage probability,
that much of the research has used techniques developed for packet error rate etc. but even till date BER is
Additive White Gaussian Noise (AWGN) channel as the assumed to be one of the most important and critical
starting point. In several cases this has put unnecessary parameter for analysis of communication systems. It
restrictions on the system design and has led to poor
performances. This means that less works use the Rayleigh
is the one that is most revealing about the behavior
fading channel as the starting point. In this paper, we present of a communication system. The search for and
a statistical frame work for estimation and calculation of ability to find a statistical frame work for problems
average bit error rate (BER) of binary frequency shift keying dealing with the evaluation of BER over generalized
(BFSK) under Rayleigh fading channels. A detailed approach fading channels motivated us to write this paper.
to obtain BER expressions for BFSK in presence of AWGN
channel is available in several published material. A little or
no material is available for fading channels from the Two different approaches have been discussed in
perspective of undergraduate students. The statistical this paper to obtain mathematical model of BER
framework presented is a step by step procedure to derive the estimation and calculation for BFSK under Rayleigh
expressions for BER. Two different approaches have been fading channel. In the first method error function is
investigated in this paper to obtain these expressions. The
expressions thus obtained have been used for comparative
used while the second approach is based on Moment
study of BPSK and BFSK in terms of BER through the Generating Function (MGF) using Gaussian Q
simulation. Following the same approach BER of other shift function. Starting from the basics of the Rayleigh
keying techniques can be easily derived and investigated. fading we develop the final expression in a lucid
Keywords —Gaussian Q function, Error function, MGF, Bit manner so that undergraduate students may get an
Error Rate, Rayleigh Fading, BFSK. exposure to the mathematics behind the expressions
of the BER over fading channels.
I. INTRODUCTION A similar work for BPSK has been presented by the
authors of this paper and is available in [4].
Regardless of the branch of science or engineering,
theoreticians have always been enamored with the
notion of expressing their results in the form of II. ERROR FUNCTION BASED APPROACH
closed-form expressions. But the elegance of closed-
form solution is often neglected. First due to
complexity of its form and the second difficulty in In a multipath channel, received signal consists of a
large number of plane waves whose complex low
computing it numerically. Under such a scenario,
COMO401-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
pass signal can be modeled as circular symmetric However in the presence of Rayleigh fading
Gaussian random variable and is of the form, amplitude , the effective bit energy to noise ratio
re jim (1) 2 Eb
is . So the conditional BEP for given value of
If none of the multipath component is dominant, N0
re and im are Gaussian processes of 0 mean is
independent and identically distributed Gaussian
random variable of 0 mean and variance 2 . 1 2 Eb 1
p(b; ) erfc( ) erfc( ) (6)
2 2 N0 2 2
The magnitude is a Rayleigh random variable
which has a probability density given as [6, page no. 2 Eb
524]. Where
N0
2
p( ) 2 exp 2 (2) To find the error probability over all random values
2 of 2 , we have to evaluate the conditional PDF
This represents the Rayleigh fading channel model p( E; ) over the PDF of .
which mitigates the signal propagation in Non Line
of Sight (NLOS) channel suffering from multipath For Rayleigh fading, is a Rayleigh distributed
fading. random variable, therefore 2 is a chi-square
When Rayleigh fading is present, the received distributed with two degrees of freedom. Since
carrier amplitude is attenuated by the fading 2 is chi-square distributed, is also chi-square
amplitude , which is a random variable with mean
distributed. The PDF of is given as
square value 2 and probability density function
(PDF) dependent on the nature of the fading 1
p ( ) e 0 (7)
channel.
Now the received instantaneous signal power is Now to find the BEP over Rayleigh fading channel,
attenuated by Eb N0 and therefore instantaneous we have to evaluate conditional BEP over PDF of
Signal to Noise Ratio (SNR) is given as [6, page no. [page no252, 7]
527]
Eb N0
2
(3) pb ( E ) P(b; ) p ( )d (8)
0
Average SNR is given as
1
2 0
Eb N 0
2
(4) Or pb ( E ) erfc ( )e d (9)
2
Due to such a condition imposed by fading, Bit Consider
Error Probability (BEP) of any modulation scheme
is obtained by replacing Eb N0 by in the
I1 erfc e d (10)
expression for AWGN performance. This is known
2
as conditional BEP and it is denoted by pb ( E; ) .
The BER of BFSK over AWGN channel is given by Now putting x and b (11)
2
[6, page no. 529] as
eq. (10) can be written as
1 Eb
pb ( E ) erfc( ) (5)
2x
2 2 N0 I1 2 erfc( x )e b dx (12)
COMO401-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Or I1 2I 2 (13)
2
0
1
pb ( E ) [ erfc( y )e erfc
2x
2 2 2
Where I 2 erfc( x )e b
dx (14)
2 x (23)
b b
erfc( x )e b
I3 (15)
2 2 This finally results into desired result as shown
b2 1 below
x
Where I3 e b 2
x dx (16)
1
b2 Pb ( E ) 1 (24)
Now making the substitution x y , eq. (16) 2
2
b
can be computed as
1
b III. MGF BASED APPROACH
b2
y 2
I3 e y dy (17)
At this stage we make use of a pre-established result The BEP of BFSK over AWGN in terms of
given as Gaussian Q function is given by [7, page 229]
1
e y dy erfc( y )( ) Eb
y
pb ( E ) Q(
2 (18) ) (25)
N0
Now substituting eq. (18) in eq. (17), we have
Therefore conditional BEP can be given as
b
I3 ( )erfc( y ) (19) p(E; ) Q( ) (26)
b2
When characterizing the performance of coherent
b (b 2) x
( )erfc (20) digital communication, the generic form of
b2 b expression for BEP involves Gaussian Q function
with an argument proportional to square root of
instantaneous SNR of received signal. In case of
Putting eq. (20) in eq. (15), we have communication over a slowly varying fading
channel, the instantaneous SNR per bit is a time-
varying random variable with a PDF P .
0
1
2
pb ( E ) [ erfc( y )e erfc
2 2
To compute the average BEP, we must evaluate an
(21) integral consisting of above mentioned Gaussian Q
function and fading PDF, that is
From eq. (13) and making us of eq. (21) we have I Q(a ) P ( )d (27)
0
COMO401-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
10
1
pb ( E ) (1 ) (35) -3
1
10
2
-4
10
IV. RESULTS
0 5 10 15 20 25 30 35
The results obtained above have been simulated Eb/No, dB
using MATLAB and following graph is obtained
which compares the performance of BFSK over
AWGN channel and Rayleigh fading channel. [Figure 2: BER of BPSK for Rayleigh fading
channel]
COMO401-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
[1] G. D. Forney, L. Brown, M. V. Eyuboglu, J. L. Moran III, [15] S. Sampei, T. Sunaga, “Rayleigh Fading Compensation
“The V.34 High-Speed Modem Standard”, IEEE Comm. for QAM in Land Mobile Radio Communications”; IEEE
Mag., Vol. 34, No. 12, pp. 28-33, Dec.1996. Trans. Vehic. Techn. Vol. 42, No. 2, pp. 137-147, May 1993.
[2] A. Acampora, “Wireless ATM: A Perspective on Issues [16] S. A. Al-Semari, T. E. Fuja, “I-Q TCM: Reliable
and Prospects”, IEEE Pers. Comm. Mag., Vol. 3, No. 4, pp. 8- Communication Over the Rayleigh Fading Channel Close to
17, Aug. 1996. the Cutoff Rate”, IEEE Trans. Inform.
Theory, Vol. 43, No. 1, pp. 250-262, Jan. 1997.
[3] Digital communication by Proakis fourth edition, chapter
5, page no. 257. [17] P. Y. Kam, H. M. Ching, “Sequence Estimation over the
Slow Nonselective Rayleigh Fading Channel with Diversity
[4] Vicky Singh, Amit Sehgal “ Error modelling of BPSK Reception and its Application to Viterbi Decoding”, IEEE J.
over Rayleigh fading channel- A Statistical Pedagogic Select. Areas Comm., Vol. 10, No. 3, pp.562-570, April 1992.
Approach”, National Conference on Advancement of
Electronics and Communication Engineering, IIMT college of [18] A. D. Polyanin and A. V. Manzhirov, Handbook of
engineering, Gr.Noida, February 2011. Integral Equations, Second Edition, Updated, Revised and
Extended, Chapman & Hall/CRC Press, 2008, 1144 pages.
[5] M.K. Simon and M.S. Alouini, “Simplified noisy
reference loss evaluation for digital communication in the
presence of slow fading and carrier phase error,” IEEE Trans.
Vet. Tech., vol. 50, no. 2, pp. 480-486, Mar. 2001] “Digital
Communication over fading channels”, alouini.
COMO401-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO401-6
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
ON PLC SYSTEM
Priyank Sharma, Student M.Tech (Comm.), Ravikant Saini, Astt. Prof., (ECE)
Shobhit University, Shobhit University,
Meerut (UP) Meerut (UP)
sh_priyank@yahoo.com
sravikantk 2@rediffmail.com
Abstract- In this paper, performance of DMT has Optimization of a transmission scheme can only be done
been studied for a PLC system. The power line has after an accurate channel model is made available.
been modeled as an access link channel. To observe the Although some models have been proposed [2-6] ,but there
effect of impulsive noise, a simulation model based on is no widely accepted model similar to mobile radio
statistical properties of impulsive noise has been used. channels. DMT as a multicarrier modulation is the best
Performance has been observed with AWGN only and technique for combating the frequency selective fading [7,
AWGN and impulsive noise both with different error 8]. Adaptivity in terms of bit loading is an added feature
correcting codes. System has been observed by 256 no against the time variation of the multipath channel. Strong
of sub carriers and Message Transfer Unit (MTU) size forward error correction strategy is required to overcome
280 bits. the hostile noise environment, which includes background
noise, colored noise and a combination of impulsive noises
Keywords Component-Power Line Channel, Access [9- 12].
Link Channel, DMT, MTU.
I. INTRODUCTION
The liberalization of telecommunications and the
deregulation of electricity utilities have added new
dimensions to the potential application of the electricity
infrastructure for the most efficient use of the local loop
[1]. The main focus is and will continue to be on the
connection between house and transformer as a solution for
the last dirty mile problem to compete with other last mile
technologies like digital subscriber line (DSL),wireless
local loop, or telephone lines. Furthermore, new interest
arises due to recent developments regarding in-house
networking.
Due to the fact that power lines were not conceived for II. CHANNEL MODELS
communication purposes but to deliver electrical energy
(which involve signals of much lower frequency and The modeling problem for PLC, which consists of
higher power), the medium characteristics are certainly investigating the characteristics of the power network as a
adverse. The channel is time and frequency variant, and communication channel , is a serious one .Power line grid
exhibits remarkable differences between locations, has tree like structure. Signal faces reflections at every
according to the network topology; kind of wire installed branching, empty socket and connected loads. Thus a
and connected loads. Even in a specific power circuit the complex echo scenario arises leading to frequency
conditions change depending on the selected transmission selective fading.
path (i.e. sockets where the transmitter and receiver are
plugged)
COMO402-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
N
H ( f ) gi .e ( a0 a1 . f e j 2fdi / v p
k
) di
.
i 1
where N is the number of the relevant multipaths,a0 and a1 2Npoint complex sequence is transformed using inverse
are attenuation parameters and k is an exponent taking fast Fourier transform (IFFT) to generate 2N point real
value in the range of 0.5 to 1; is the path length and vp is samples for transmission through the channel. In order to
the speed of EM wave in the material. mitigate ISI and maintain the continuity between
consecutive DMT symbols, the last 25% of each 2N-
III DISCRETE MULTITONE sample block are circularly wrapped to prefix the block
itself. After receiving the transmitted signal, the receiver
DMT efficiently divides an input serial data bit stream discards the prefixed samples; then the remaining received
into a large number of low-data-rate streams and assigns samples are demodulated into complex symbols by the 2N
them to a set of parallel and orthogonal subcarriers point FFT. These complex symbols are decoded using a
according to the SNR of respective subcarriers [8]. SNR of QAM decoder. The in phase and quadrature components of
each subcarrier is estimated in an initial training session. the complex symbols are used by the QAM decoder to
Fig 1 shows the block diagram representation of a basic generate the received data per subcarrier. Finally a parallel
DMT transceiver. to serial conversion recovers the original transmitted data
bit stream. For the better results different algorithm may be
The data bits assigned to each subcarrier are mapped onto
used[13-16]. The vital components of a DMT transceiver
a Quadrature Amplitude Modulation (QAM) constellation
system are the IFFT and FFT at the transmitter and
to form a complex sample, and then the resulting N
receiver , respectively.
complex samples from the N subcarriers are extended to be
a 2N point complex –conjugate symmetric sequence. This
COMO402-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Performance of DMT is evaluated by below methods: Figure (3) 256_280 With AWGN & Impulsive Noise
In order to study the performance degradation due to Figure(4)128_112 With AWGN Only
impulsive noise, two noise environments have been
considered.
1. AWGN only
2. AWGN and Impulsive noise.
Fig.(4) & Fig(5) shows the Result 128_112 AWGN & V. CONCLUSION
Impulsive noise
Performance of DMT system has been evaluated for the
Access link. In impulsive noise analysis the error rate is
higher than the AWGN environment as expected. This is
because of the burst errors caused by Impulsive noise. In
order to improve the performance further higher error
correcting codes can be used.
REFERENCES:
COMO402-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
IEEE Trans. Electromagn. Compat.,vol. 40, pp. 331–336, [10] M.Zimmermann and K.Dostert, “Analysis and
1998. modeling of impulsive noise in broadband powerline
communications,”IEEE Trans Electomagn. Compat.,
[3] H. Meng, S. Chen, ”Modeling of Transfer vol.44,No.1,pp.249-258,Feb.2002.
Characteristics for The Broadband Power Line [11]R. M. Vines, M. J. Trussel, L. J. Gales, and J. B.
communication channel”, IEEE Transactions on O’Neal, Jr., “Noise on residential power distribution
Power Delivery, Vol. 19, No 3, pp. 529-551. Jul. 2004. circuits,” IEEE Trans. Electromagn.Compat., vol.
[4]L.T.Tang,P.L.So,E.Gunawan,Y,L.Guan,S.Chenand EMC-26, pp. 161-168, Nov. 1984
T.T.Lee,”Characterization and modeling of In-Buiding [12] D. Anastasiadou and T. Antonakoupoulos,
Power Lines for High speed Data Transmission,”IEEE ”Multipath characterization of Power-Line Networks”,
Trans.Power Delivery,vol-18,No-1,pp.31-40,Jan.2003. IEEE Trans. On Power Delivery, vol. 20,pp. 90-99, 2005.
[5] M. Zimmermann and K. Dostert. "A Multipath Model [13] E.Guerrini,G.Dell’ Amico,P.Bisagila, L.Guerrieri,”
for the Powerline Channel,'' IEEE Trans. Commun.. bit-loading algorithm and SNR estimate for Homeplug
vol.50. no. 4. Apr. 2002, pp. 553-59. AV” IEEE ISPLC’07,pp419-424,26-28 March 2007.
[6] A. Mori, Y. Watanabe and M. Tokuda & K. [14]F.S.Muhammad, J.-Y.Baudais, J.-F.Helard and
Kawamoto “Transmission Characteristics of an OFDM M.Crussiere,” Coded adaptive linear precoded discrete
Signal for Power Line Communication System with High multitone over PLC Channel,” ISPLC’08,April 2008
Bit Rate”.IEEE Communication Magazine, July 2007. [15] A.M.Wyglinski,F.Labeau and P.Kabal,”Bit loading
[7] E. Biglieri. Coding and modulation for a horrible with BER constraint for multicarrier systems,” IEEE
channel. IEEE Communications Magazine, 41:92–98, Trans. On wireless comm.., vol.4,pp. 1383-1387,july
May 2003. 2005
[8] S. Baig and N. D. Gohar. A discrete multitone [16] M. H. L. Chan and R.W. Donaldson, “Amplitude,
Transceiver at the heart of the PHY layer of an in-home width, and interarrival distributions for noise impulses on
power line communication local area network. IEEE intrabuilding power line 3, pp. 320–323, Aug. 1989.
Communications Magazine, 41:48–53, Apr. 2003.
[9] M. Zimmermann and K. Dostert. An analysis of the
broadband noise scenario in power line networks. In IEEE
International Symposium on Power Line Communications
and Its Applications ISPLC ’2000, pages 131–138, 2000.
COMO402-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Design and Analysis of Microstrip Antennas for W-LAN and Ku-band Applications
1
Ravindra Kumar Yadav and 2Ram Lal Yadava
1
Department of Electronics and Communication Engineering,
I.T.S. Engineering College, Greater Noida, Uttar Pradesh, India
2
Department of Electronics and Communication Engineering,
Galgotia’s College of Engineering & Technology, Greater Noida, Uttar Pradesh
ravipusad@gmail.com
Abstract
This paper deals with the design analysis methods for microstrip antennas (MSA) meant for W-LAN and Ku
Frequency band applications. The design analysis methods for MSAs can be broadly classified into two
groups. In the first group, the methods are based on equivalent magnetic current distribution around the
patch edges; The Transmission Line Model and the MNM. The second group methods are based on the
electric current distribution on the ground plane (similar to dipole antennas), and the popular methods are
Method of Moments, Finite Element method, Spectral Domain Technique and Finite-difference Time
Domain method. In this work, both the techniques have been implemented to analyze the fabrication and
performance of the antenna. The Transmission Line Model and the Method of Moments has been used.
Three antennas structure have been considered namely; proximity and aperture coupled, and slot coupled
MSAs.
Keywords: Microstrip antenna, Ultra wideband coupled patch antenna and circularly polarized MSA.
I. INTRODUCTION II .THEORY
Since there has been an increase in the demand of Microstrip antennas are used due to its
Microstrip Antennas, these are manufactured in a several advantages like small size, ease of
large scale, but there are limitations with the design fabrication, low cost and many others. Polarization
of these antennas. This is because proper analysis is is an important part of the study of microstrip
required before fabrication and then simulation for antennas. Usually the polarization of an antenna is
characteristics investigations [1-5]. Therefore defined as the wave radiated by the antenna in a
present paper aims to describe the design analysis of particular direction. This is usually depends on the
Microstrip Antennas using the Transmission line feeding technique [4-5]. However in general,
model and the Method of Moments and study the orientations of the electric field vector of a plane
performance characteristics as well. For this, IE3D wave relative to direction of propagation define the
software is used, and since the transmission line polarization of wave. If an electric field vector of an
model is an approximate model, genetic algorithm EM wave is parallel to x-axis, wave is said to
is also being implemented to optimize the design. linearly x-polarized. This wave could be generated
The Method of Moment model is used to determine by a straight wire antenna parallel to x-axis.
the current distributions and field distributions of Similarly y-polarized waves can also be defined and
the antenna. An aperture coupled microstrip antenna generated. If two plane waves of equal amplitude
is chosen for the design using transmission line and orthogonally polarized are combined with
model and a wire antenna is chosen for the Method phase – difference, the resulting wave is circularly
of Moments model [6-10]. polarized and electric field vector describe a circle
centered on propagation vector. The field vector
will rotate by for every wave length traveled.
COMO421-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Circular polarization generated as either RHCP or opportunity to reduce feed line radiation
LHCP. Right hand circularly polarized wave while maintaining a relatively thick rotation.
describes a wave with E-field vector rotating The main advantage of Circular Polarization
clockwise when looking in the direction of is that regardless of substrate for the
propagation [7]. Bandwidth: The band width of radiation patch. The input impedance of the
antenna is usually defined as the frequency range antenna is affected by the overlap of the
within the performance of the antenna with respect patch and the feed line, and by the substrate.
to a certain characteristics. It is expressed as the This feature adds degree of freedom in the
percentage of the difference between upper and design, but may complicate the task of
lower frequency to the center frequency; selecting an optimum design.
fu f
BW 100% .
fo
Bandwidth of an antenna can also be defined in
terms of radiation patterns or VSWR/reflected
VSWR 1
power as follows; BW .
Q VSWR
Where, Q is a quality factor. Feeding techniques:
Popular feeding techniques are the coaxial probe Fig 2. Geometry of proximity coupled MSA.
feed, microstrip line feed, aperture coupled and the
proximity fed antenna. 3rd antenna analyzed is a rectangular slot
antenna for ultra wideband applications. It is
Types of antennas analyzed: shown below (Figs 3 and 4).
Fig 1. Geometry of aperture coupled MSA. Fig 4. Geometry of slot coupled MSA (Side view).
COMO421-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Brief introduction of methodologies used: ultra wideband applications and phased array
systems are shown below in table 2.
IE3D is simulation software used to
simulate various geometries of antennas. Table 2. VSWR, Gain and BW
It is commercially available software.
For design optimization process genetic W1 W2 L1 L2 L3 VSWR Gain BW
mm mm mm mm mm MHz
algorithms are used which work on the
theory of natural reproduction process. 44 36 16 5 4 1.0382 3.191 690
For this optimization process a fitness
function is programmed using the
transmission line model of the geometry. The overall performance of the simulation software
Method of Moments is another method (IE3D) for the aperture coupled microstrip antenna
which gives idea about the current and (substrate thicknesses: 1.2 mm (lower), 1.5748 mm
field distribution with respect to (upper),) (dielectric constant: 1.9 lower, 3.9 upper is
normalized wavelength. listed in Table below in table 3.
IV. CONCLUSIONS
The antenna geometry was again designed with a
substrate provided and the results are; Both transmission line model and the
method of moment’s model have been
Ls Ws VSWR Bandwidth programmed. In case of transmission line
mm mm MHz model, we observe that by varying the
2.828 1.414 1.0275 223 length, width and other dimensions of the
patch, we can vary parameters like VSWR
and Frequency of operation.
Therefore there is increase in bandwidth. The In case of method of moment’s model we
simulation results for the Rectangular slot antenna can get the detailed explanation of current
(resonating at 2.44 GHz) with patch substrate for and field distribution provided we have the
COMO421-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
Green’s function available for that antenna Proximity however rectangular patch shows
structure. 2.4 GHz. The optimized results of VSWR
Aperture coupled microstrip antenna has a vary from 1.064 to 1.824.
higher gain than the rectangular slot antenna
with patch stub but the latter has a higher Acknowledgment
bandwidth. The proximity coupled antenna
showed very lee gain but a high bandwidth. The authors are very thankful to Prof. K.V.
Genetic Algorithms are very useful to Karthikayen, IIT-R, for his motivation and
implement the transmission line model when support to complete this research work in
we use the VSWR as the fitness function. time.
COMO421-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO422-1
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO422-2
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO422-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO422-4
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
IV. COMPARISON TABLE [6] L.C.Tran, J. Sebem, B. J. Wysocki, T. A. Wysocki, T. Xia and Ying
Zhao, “Complex Orthogonal Sequences from Amicable Hadamard
In the comparison table the algorithms for generating one Matrices,” Vehicular Technology IEEE Conference, Vol-3, pp 1490-
dimensional optical orthogonal codes has been compared on 1493, 2004.
the basis auto-correlation and cross correlation constraints [7] Abrham, J. Kotzig, “Skolem Sequences and Additive Permutations,”
Book Discrete Math, pp 143–146, 1981.
a , c and number of codes generated. [8] Manoj Choudhary, P. K. Chatterjee and Joseph John, “Optical
Orthogonal Codes using Quadratic Residues,” in Proc. of National
Conference on Communication, IIT Kanpur, 2001.
Auto- Cross- [9] R.C.S.Chauhan, Rachna Asthana and Y.N.Singh, “Uni-polar Orthogonal
Correlati Correlat Code No. of Codes : Design, Analysis and Applications,” HiPC 2010.
Type of Weight
on ion length codes
algorithm (w)
( a ) ( c ) (n) generated
Prime
p-1 2 p2 p p
sequence
Quasi prime Q-1 2r qp q p
Quadratic
2 4 p2 p p-1
congruence
Hadamard
t-1 T 4t - 1 2t - 1 <n
matrix
Skolem
1 1 6M+1 3 M
sequence
Quadratic
3 p 1
2
p 2
p p-1
residues
Table of
p-1 p–1 p2 p p-1 p-1
prime
Proposed user Johnson‟s
w-1 w-1 w<<n
algorithm defined bound
q varies from (r-1) p<q< rp where r, q is the positive integer,
p is a prime number, m is an integer, M is Skolem sequence
order
V. CONCLUSION
The Optical CDMA system employing optical orthogonal
codes as the optical signature sequence to the OCDMA users.
One dimensional optical orthogonal codes are designed on the
basis of some algorithms found in literatures [ 3-5, 7-9 ] using
MATLAB based programs along with the calculations of auto-
correlation and cross correlation constraints a , c
respectively. In this paper the algorithms of generating one
dimensional optical orthogonal have been compared on the
basis of some parameter like code length „n‟, code weight „w‟,
and auto-correlation and cross correlation constraints a , c
which are responsible for BER performance of OCDMA
system. In future these coding schemes can be improved for
generating the groups of one dimensional optical orthogonal
codes with maximum possible codes.
REFERENCES
[1] A. Stok and E. H. Sargent, “The Role of Optical CDMA in Access
Networks,” IEEE Commun. Mag., vol. 40, no. 9, pp. 83–87, 2002.
[2] Fan R.K. Chung, Jawad A. Salehi and Victor K. Wie, “Optical
Orthogonal Codes: Design, Analysis, and Application,” IEEE Trans.
Information Theory, vol. 35, no. 3, may 1989.
[3] A.A.Shaar and P.A.Davis, “Prime sequences and Quasi-optimal
sequences for channel code division multiplexing,” Electronics Letters,
vol. 19, pp. 888-889,October 1983.
[4] Andrew S. Holmes and Richard R.A. Syms, “All Optical CDMA using
Quasi Prime Codes,” IEEE Journal of lightwave technology, vol. 10, no.
2, Feb 1992.
[5] M. Srinivasan, G. Geetha, M. Meenakshi, “Comparision of Prime Code,
Extended Prime Code and Quadratic Congruence Code using a
Normalized Throughpu Metric,” ICTON, pp.168-176, 2004.
COMO422-5
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
DECISION n=1 n=2 n=3 n=4 n=5 n=6 n=7 n=8 n=10
P_e_hard 0.0239 0.8744 2.903 5.2896 7.5817 9.6382 11.4406 12.034 15.3502
P_e_soft 0.0000 0.0047 0.0623 0.2333 0.5246 0.9107 1.3607 1.8489 2.4600
extrinsic messages about the same symbols. error in soft decoding and hard decoding
The extrinsic message is generated by scheme are shown in table in table.2
considering the a priori constraints that exist
between the input and output sequences of Table- 2
the encoder. In this section we will give more
precise definitions of the “message” and on
the input output relationships of a SISO
block. Moreover we will show efficient
algorithms to perform SISO for some special
types of encoders.
A. Definition of the input and output metrics
A SISO module generally works
associated to a known mapping (encoding)
between input and output alphabets [4].
A SISO module is a four port device
that accepts some messages of the input and
output symbols of the encoder and provides
homologous output extrinsic messages. We
will consider the following two types of
normalized messages:
L(x) represents the ratio between the The probability of error is low and higher
likelihood of the symbol being 1 and 0 "_ decoding throughput in soft decoding
scheme then hard decoding scheme as
shown in fig.-1
COMO431-3
CONFERENCE ON “SIGNAL PROCESSING AND REAL TIME OPERATING SYSTEM (SPRTOS)” MARCH 26-27 2011
COMO431-4