PAPR ANALYSIS AND SIMULATION FOR 3GPP LTE SYSTEM

A B.Sc Engineering Thesis

by

S.M. Mahmud Hasan Roll No: 074019

Department of Electronics and Telecommunication Engineering RAJSHAHI UNIVERSITY OF ENGINEERING & TECHNOLOGY September 2012

PAPR ANALYSIS AND SIMULATION FOR 3GPP LTE

by S.M. Mahmud Hasan Roll No.: 074019

A thesis submitted in partial fulfillment of the requirements for the degree of

BACHELOR OF SCIENCE in Electronics and Telecommunication Engineering

to the

Department of Electronics and Telecommunication Engineering

RAJSHAHI UNIVERSITY OF ENGINEERING & TECHNOLOGY

September 2012
1

Declaration

This is to certify that the thesis work PAPR Analysis and Simulation for 3GPP LTE System by S.M. Mahmud Hasan, bearing Roll no. 074019 has been carried out under my supervision as a requirement for the degree of Bachelor of Science in Electronics and Telecommunication Engineering.

Md. Munjure Mowla Lecturer Department of Electronics and Telecommunication Engineering Rajshahi University of Engineering & Technology Rajshahi - 6204.

Acknowledgement

On the submission of my thesis report of PAPR Analysis and Simulation for 3GPP LTE System, I would like to extend my gratitude and sincere thanks to my supervisor, Md. Munjure Mowla, Lecturer, Department of Electronics and Telecommunication Engineering for his constant inspiration and support during the course of my work in the last one year. I truly appreciate and value his esteemed guidance and encouragement during execution of thesis work from the beginning till end of this thesis. He has been great sources of inspiration to me and I thank him also for imparting me immense knowledge in the field of communication which made my work a lot easier. I am indebted to his for having helped me in taking various problem statements and providing methods and techniques for the solution of it. This thesis would have been difficult to accomplish without his continuous moral support.

S.M. Mahmud Hasan Roll No.- 074019 RUET, Rajshahi. September 09, 2012.

Abstract

The highest bit rates in commercially deployed wireless systems are achieved by means of Orthogonal Frequency Division Multiplexing (OFDM). The next advance in cellular systems, under investigation by Third Generation Partnership Project (3GPP), also anticipates the adoption of OFDMA to achieve high data rates. But a modified form of OFDMA i.e. SCFDMA (Single Carrier FDMA) having similar throughput performance and essentially the same complexity has been implemented as it has an edge over OFDMA having lower PAPR (peak to average power ratio). SCFDMA is currently a strong candidate for the uplink multiple access in the Long Term Evolution of cellular systems under consideration by the 3GPP. In the thesis, Peak to Average Power Ratio (PAPR) analysis of OFDMA & SCFDMA with different subcarrier mapping has been performed. Though SCFDMA had larger ISI it has lower PAPR which help in avoiding the need of an efficient linear power amplifier. Various modulation techniques and various parameters have been changed to compare the PAPR for OFDMA & SCFDMA. Many techniques have been studied for reducing the PAPR of a transmitted OFDM signal. In general, in LTE the cost and complexity of generating the OFDM signal with acceptable Error Vector Magnitude (EVM) is left to the eNodeB implementation. As OFDM is not used for the LTE uplink, such considerations do not directly apply to the transmitter in the UE. Techniques for PAPR reduction of OFDM signals can be broadly categorized into three main concepts: Clipping and Filtering, Selected Mapping and Pre-coding Technique.

Clipping & Filtering technique has been introduced for PAPR reduction of OFDM signals. The effects of high power amplifier and the channel noise on the OFDM signals have been also analyzed and then introduced clipping & filtering as a PAPR reduction method to reduce this effect. This technique consists of oversampling the original signal by padding the input signal with zeros and processing it using a longer IFFT. The oversampled signal is clipped and then ltered to reduce the out-of-band radiation.

Contents
Declaration Acknowledgement Abstract Contents List of Tables List of Figures Acronyms 2 3 4 5 10 10 13

CHAPTER 1:
1.1 1.2 1.3
1.4

Introduction
16 17 18
19

Introduction 3rd Generation Partnership Project LTE in the Mobile Radio Landscape
Evolution of 4G

1.5 1.6 1.7 1.8

Requirements for Long Term Evolution Multi Carrier Modulations Objective of Thesis Scope of Thesis

21 22 22 22

Chapter 2:
2.1 2.2 2.2.1 2.2.1.1 2.2.2

LTE Network Architecture

Introduction Overall Architectural Overview The Core Network Non-Access Stratum (NAS) Procedures The Access Network 5 24 24 26 28 29

2.3 2.4 2.5 2.6 2.6.1 2.6.2

Roaming Architecture Inter-Working with other Networks Inter-Radio Access Technologies (RAT) Mobility Connected Mode Inter-RAT Mobility Handover to LTE Mobility from LTE

31 32 33 34 34 34

Chapter 3:
3.1 3.2 3.3 3.3.1 3.3.2 3.4 3.5 3.6 3.7 3.8 3.9

Physical Layer in the LTE Uplink

Introduction LTE Uplink Requirements SC-FDMA Principles SC-FDMA Transmission Structure Time-Domain Signal Generation SC-FDMA Frame Structure Uplink SC-FDMA Parameters Modulation Implementation of the SC-FDMA Transceiver LTE Uplink physical channels LTE uplink transport channels 36 36 37 37 37 38 39 40 40 41 41

Chapter 4:
4.1 4.2 4.3 4.3.1 4.3.2

Physical layer in the LTE downlink

Introduction LTE Downlink Requirements OFDM Principles Orthogonal Multiplexing Principle Importance of Orthogonality 42 42 43 43 46

4.3.3 4.4 4.5 4.6 4.7 4.7.1 4.7.2 4.8 4.8.1 4.8.2 4.8.3 4.9

Guard Interval OFDM Frame Structure Downlink OFDM Parameters Mapping of Subcarriers Implementation of the OFDM Transceiver Binary Source Generator Modulation Downlink Data Transmission Modulation Downlink Reference Signal Structure Cell Search Latency Requirement

48 48 49 50 51 51 51 52 52 52 54 54

Chapter 5:
5.1 5.2 5.2.1 5.3 5.4 5.4.1 5.5 5.6 5.7 5.8 5.8.1 5.8.2

PAPR Calculation for SCFDMA & OFDMA

Introduction SCFDMA Block Diagram of SC-FDMA OFDM OFDMA Block Diagram of OFDMA Description of Problem Statement Mathematical Calculation for PAPR Comparison of PAPR for OFDMA and SCFDMA Significance of Pulse Shaping Filter in PAPR Analysis Sinc Filter Raised Cosine Filter 7 55 55 57 58 59 61 62 64 65 65 66 67

5.8.3 5.9 5.9.1 5.9.2 5.9.3

Gaussian Filter PAPR Reduction Techniques for OFDM signal Clipping and Filtering Selective Mapping Pre-coding Technique

68 68 69 71 72

Chapter 6:
6.1 6.2 6.3 6.4 6.5 6.6

Characteristics of Mobile Radio Channel

Introduction Types of Fading Small-scale Fading Critical Channel Parameters Types of Small-scale Fading Rayleigh and Ricean Distribution 73 73 74 74 75 76

Chapter 7
7.1 7.2 7.3 7.4 7.5 7.5.1 7.5.2 7.6

Channel Estimation in OFDM

Introduction Block type of Pilot Arrangement Comb type of Pilot Arrangement Working Environment Mathematical Analysis of the Channel Estimators Least Square Error (LS) Estimation Minimum Mean Square Error (MMSE) Estimation Modified MMSE Estimation 78 79 79 79 80 81 82 83

Chapter 8:
8.1 8.2 8.3

Simulations & Results

OFDM Signal and its spectrum with Guard Interval Comparison of PAPR for OFDMA and SCFDMA Investigation of Clipping & Filtering method as PAPR Reduction Technique for OFDM signals 92 84 85

Conclusion & Future Scope Reference Appendix A Appendix B Appendix C

99 100 102 107 113

List of Tables
Page No.
39 50

Table No.
3.1 4.1 4.2

Name of Table
Uplink parameters for SC-FDMA transmission Downlink parameters for OFDM transmission Normalization factor for M-QAM modulation Schemes in E-UTRA downlink

52

List of Figures

Fig No.
1.1 1.2

Name of Figure
Radio Access Network Milestones Approximate timeline of the mobile communications standards Landscape

Page No.
17

19 25 27 30

2.1 2.2 2.3 2.4

The EPS network elements Functional split between E-UTRAN and EPC Overall E-UTRAN architecture Roaming architecture for 3GPP accesses with P-GW in home Network

32 33 34 35 38 38 39 40 40 44 44 45 45 48

2.5 2.6 2.7 3.1 3.2 3.3 3.4 3.5 4.1 4.2 4.3 4.4 4.5

Architecture for 3G UMTS interworking Uplink S1 CDMA2000 tunneling procedure. Mobility from LTE SC-FDMA time-domain transmit processing Generic frame structure (TDD or FDD) Slot structure Block diagram of the SC-FDMA transmitter in LTE Block diagram of the SC-FDMA receiver in LTE Serial-to-parallel conversion operations for OFDM OFDM Transmitter OFDM receiver OFDM cyclic prex insertion Insertion of cyclic prefix

10

4.6

OFDM Frame structure in LTE. A radio frame is divided Into 20 slots of 0.5 ms each having 6 or 7 OFDM symbols 49 50 51 51

4.7 4.8 4.9 4.10

Placement of occupied subcarriers Block diagram of the OFDM transmitter in LTE Block diagram of the OFDM receiver in LTE The reference symbol structure for one slot with 6 OFDM Symbols using two antennas

53

5.1

Difference between channel representations between OFDMA And SCFDMA 56 57

5.2 5.3

Tx and Rx structure of SCFDMA (M > N) Spectral efficiency of OFDM compared to classical multicarrier Modulation: (a) Classical multicarrier system spectrum (b) OFDM system spectrum

58 58 60 60 61 62

5.4 5.5 5.6 5.7 5.8

Difference between OFDM and OFDMA Sensitivity of OFDM subcarriers with Carrier OFDM transmission spectrum Block Diagram of OFDMA Sub-carrier mapping for 3 users, 12 sub-carriers and 4 sub-carriers Per user

63 65 66 67 68 69

5.9 5.10 5.11 5.12 5.13 5.14

PAPR distribution for different numbers of OFDM subcarriers The Transfer Function of Sinc Filter The Transfer Function of Raised Cosine Filter The Transfer Function of Gaussian Filter Simplified clipping and filtering with Optimum value of The clipping and frequency domain filtering of the input OFDM Signal

70

5.15

Block diagram of SFBC-OFDM transmitter with two transmitters Antennas and the selective mapping (SLM) method for PAPR Reduction 71

5.16

Block diagram of pre-coding technique for PAPR reduction of OFDM signal 72 77 79 80 81 82

6.1 7.1 7.2 7.3 7.4

Rayleigh fading channel with two path sine wave input Two basic types of pilot arrangement for OFDM channel estimation General estimator structure SNR vs BER using LSE estimator for an OFDM channel SNR vs MSE for an OFDM system with MMSE / LSE estimator

11

7.5 8.1

SNR vs SER for an OFDM system with MMSE / LSE estimator OFDM signal and its spectrum with Guard Interval (Graph on time domain)

83

84

8.2

OFDM signal and its spectrum with Guard Interval (Graph on frequency domain) 84

8.3

CCDF of PAPR for OFDMA & SCFDMA (N=64, M=512) With QPSK Modulation 85

8.4

CCDF of PAPR for OFDMA & SCFDMA (N=64, M=256) With QPSK Modulation 86

8.5

CCDF of PAPR for OFDMA & SCFDMA (N=64, M=128) With QPSK Modulation 87

8.6

CCDF of PAPR for OFDMA & SCFDMA (N=64, M=512) With 16-QAM Modulation 88

8.7

CCDF of PAPR for OFDMA & SCFDMA (N=64, M=256) With 16-QAM Modulation 89

8.8

CCDF of PAPR for OFDMA & SCFDMA (N=64, M=128) With 16-QAM Modulation 90

8.9

CCDF of PAPR for OFDMA & SCFDMA (N=16, M=128) With 16-QAM Modulation 91 92 93 94 94 95 95

8.10 8.11 8.12 8.13 8.14 8.15 8.16

Transmitted Data Phase Representation The representation of the modulated signal (QPSK) Unclipped OFDM signal Clipped OFDM signal Unclipped OFDM signal after passing through H.P.A Clipped OFDM signal after passing through H.P.A Comparison between Transmitted Data Phase Representation & Received unclipped OFDM signal

96

8.17

Comparison between Transmitted Data Phase Representation & Received clipped OFDM signal 97

12

Acronyms
1G 2G 3G 4G 3GPP 3GPP2 AMPS APN ARIB ATIS AWGN BER BPSK CCDF CCO CCSA CDMA CT CP DPSK ECM-IDLE EDGE eNodeB EPC EPS ETSI E-UTRA FDM FFT GPRS GSM First Generation Second Generation Third Generations Fourth Generations 3rd Generation Partnership Project 3rd Generation Partnership Project 2 Analogue Mobile Phone System Access Point Name Association of Radio Industries and Businesses Alliance for Telecommunications Industry Solutions Additive White Gaussian Noise Bit Error Rate Binary Phase Shift Keying Complementary Cumulative Density Function Cell Change Order, China Communications Standards Association Code Division Multiple Access Core Network & Terminals cyclic prefix Differential Phase Shift Keying EPS Connection Management IDLE Enhanced Data rates for GSM Evolution evolved NodeB Evolved Packet Core Evolved Packet System European Telecommunications Standards Institute Evolved UMTS Terrestrial Radio Access Frequency Division Multiplexing Fast Fourier Transform General Packet Radio Service Global System for Mobile communications 13

GERAN HLR HSPA HSS IDFT ITU ITU-R IMT IP ISI IMS LSE LMMSE LTE MME MMSE Mod MMSE MSE NAS NACC OS OFDM OFDMA PAPR PCRF PSK P-GW PLMN PMIP PRS PSCH QAM QoS QPSK

GSM EDGE Radio Access Networks Home Location Register High Speed Packet Access Home Subscriber Server Inverse Discrete Fourier Transform International Telecommunication Union ITU Radio communication sector International Mobile Telecommunications Internet Protocol Inter Symbol Interference IP Multimedia Subsystem Least Square Estimation Minimum Mean Square Estimation Long Term Evolution Mobility Management Entity Minimum Mean Square Estimation Modified Minimum Mean Square Estimation Mean Square Error Non-Access Stratum Network Assisted Cell Change Orthogonal Sequence Orthogonal Frequency Division Multiplexing Orthogonal Frequency Division Multiple Access Peak to Average Power Ratio Policy Control and Charging Rules Function Phase Shift Keying PDN Gateway Public Land Mobile Network Proxy Mobile Internet Protocol Pseudo-random Sequence Primary Synchronization Channel Quadrature Amplitude Modulation Quality-of-Service Quadrature Phase Shift Keying 14

RAN RAT SA SINR SAE S-GW S-TMSI SC-FDMA SER SNR SSCH TTA TTC TSG TDMA TD-SCDMA TFT TTI TDD UTRA UTRAN UMTS VoIP WAN WCDMA WiMAX

Radio Access Networks Radio Access Technologies Service & Systems Aspects Signal-to-Interference plus Noise Ratio System Architecture Evolution Serving Gateway SAE-Temporary Mobile Subscriber Identity Single Carrier-Frequency Division Multiple Access Symbol Error Rate Signal to Noise Ratio Secondary Synchronization Channel Telecommunications Technology Association Telecommunications Technology Committee Technical Specification Groups Time Division Multiple Access Time Division Synchronous Code Division Multiple Access Traffic Flow Templates Transmission Time Interval Time Division Duplex Universal Terrestrial Radio Access Universal Terrestrial Radio Access Network Universal Mobile Telecommunications System Voice-over-IP Wide Area Network Wideband Code Division Multiple Access Worldwide interoperability for Microwave Access

15

Chapter 1 Introduction

1.1 Introduction:
LTE (Long Term Evolution), marketed as 3.9G LTE, is a standard for wireless communication of high-speed data for mobile phones and data terminals. It is based on the GSM/EDGE and UMTS/HSPA network technologies, increasing the capacity and speed using new modulation techniques. The standard is developed by the 3GPP (3rd Generation Partnership Project) and is specified in its Release 8 document series, with minor enhancements described in Release 9 [1]. LTE is a wireless broadband technology designed to support roaming Internet access via cell phones and handheld devices. Because LTE offers significant improvements over older cellular communication standards, some refer to it as a 4G (fourth generation) technology along with WiMAX [2]. is considered by many to be the obvious successor to the current generation of UMTS 3G technology, which is based upon WCDMA, HSDPA, HSUPA, and HSPA. LTE is not a replacement for UMTS in the way that UMTS was a replacement for GSM, but rather an update to the UMTS technology that will enable it to provide significantly faster data rates for both uploading and downloading [3]. It is anticipated to become the first truly global mobile phone standard, although the use of different frequency bands in different countries will mean that only multi-band phones will be able to utilize LTE in all countries where it is supported. Although marketed as a 4G wireless service, LTE as specified in the 3GPP Release 8 and 9 document series does not satisfy the technical requirements the 3GPP consortium has adopted for its new standard generation, and which are set forth by the ITU-R organization in its IMTAdvanced specification [1].

16

1.2 3rd Generation Partnership Project:

The 3rd Generation Partnership Project (3GPP) unites Six telecommunications standard development organizations (ARIB, ATIS, CCSA, ETSI, TTA, TTC), known as Organizational Partners and provides their members with a stable environment to produce the highly successful Reports and Specifications that define 3GPP technologies. The Four Technical Specification Groups (TSG) in 3GPP are Radio Access Networks (RAN), Service & Systems Aspects (SA), Core Network & Terminals (CT) and GSM EDGE Radio Access Networks (GERAN).

1999 2000 01

02

03

04

Timeline (Year) 05 06 07 08

09

10

11

2012

Release 99

W-CDMA 1.28Mbps TDD HSDPA HSUPA, MBMS

HSPA+ (MIMO, HOM, etc )

Release 4

Release 5

Release 6

Release 7

Release 8

LTE LTE enhancements LTE A

Further LTE enhancements

Release 9

Release 10

Release 11+

Fig 1.1: Radio Access Network Milestones.

17

Each of the four TSGs has a set of Working Groups, which meet regularly four to six times a year. Each TSG has its own quarterly plenary meeting where the work from its WGs is presented for information, discussion and approval. Each TSG has a particular area of responsibility for the Reports and Specifications within its own Terms of Reference. 3GPP Technical Specification Group RAN, like other TSGs, ensures that systems based on 3GPP specifications are capable of rapid development and deployment with the provision of global roaming of equipment. Some of the headline 3GPP radio technologies and systems over the recent Releases [4] have been shown in the above Fig 1.1.

1.3 LTE in the Mobile Radio Landscape:

The complementary functions of the regulatory authorities and the standardization organizations can be summarized broadly by the following relationship [5] : Aggregated Data Rate = Bandwidth (Regulation & Licenses) Spectral efficiency (Technology & Standards)

From the technology and standards angle, there are currently three main organizations responsible for developing the standards meeting IMT requirements, and which are continuing to shape the landscape of mobile radio systems, as shown in Fig 1.2.

18

Approximate Timeline: 1995 2000 2010 2015

2G

3G

4G

GSM GPRS

EDGE EDGE

TD-SCDMA (China) UMTS HSDPA HSUPA HSPA+R7 HSPA+ R8

LTE
3GPP

TDD FDD

LTE advanced

TDMA/ FDMA CDMA OFDM

802.16 2004 fixed WiMAX IEEE 802.16 e mobile WiMAX

802.16 m

IS - 95 3GPP2

CDMA 2000

CDMA EVDO

CDMA EVDO Rev A

CDMA EVDO Rev B

UMB

Fig 1.2: Approximate timeline of the mobile communications standards landscape.

1.4 Evolution of 4G:

The evolution of 4G from 1G is described below [6]-[10]: 1G (early 1980s): - Analog speech communication - Analog FDMA/FDD - Ex-AMPS standard by Bell Labs 19

2G (early 1990s): - Digital speech communication - Handoff, more secure communication - TDMA and CDMA schemes - Ex-Four major standards - GSM - IS-136/IS-54 NADC, PDC(Japan) - IS-95 cdmaOne

2.5G (mid 1990s): - Improvement of data rate - Up-gradation of 2G - Ex-HSCSD, GPRS, EDGE (from GSM) IS-95B (from cdmaOne)

3G (late 1990s): - A global standard for communication - High data rate - Ex-WCDMA (UMTS), cdma2000, TD-SCDMA

4G (mid 2000s): - Based on an all-IP packet switched network. - Peak data rates of up to approximately 100 Mbit/s for high mobility such as mobile access and up to approximately 1 Gbit/s for low mobility such as nomadic/local wireless access. - Dynamically share and use the network resources to support more simultaneous users per cell. - Scalable channel bandwidth 520 MHz, optionally up to 40 MHz. - Peak link spectral efficiency of 15 bit/s/Hz in the downlink, and 6.75 bit/s/Hz in the uplink (meaning that 1 Gbit/s in the downlink should be possible over less than 67 MHz bandwidth). - System spectral efficiency of up to 3 bit/s/Hz/cell in the downlink and 2.25 bit/s/Hz/cell for indoor usage. - Smooth handovers across heterogeneous networks. - Ability to offer high quality of service for next generation multimedia support. 20

- Ex- LTE Advanced standardized by the 3GPP and 802.16m standardized by the IEEE (i.e - WiMAX)

1.5 Requirements for Long Term Evolution:

The requirements for LTE were re-defined and crystallized, being nalized in June 2005. They can be summarized as follows [5]: Reduced delays, in terms of both connection establishment and transmission latency. Increased user data rates. Increased cell-edge bit-rate, for uniformity of service provision. Reduced cost per bit, implying improved spectral efficiency. Greater exibility of spectrum usage, in both new and pre-existing bands. Simplified network architecture. Seamless mobility, including between different radio-access technologies. Reasonable power consumption for the mobile terminal.

The 3GPP LTE (Long Term Evolution) was a recent standard introduced by 3GPP group which promises high-speed data, multimedia unicast and multimedia broadcast services. The Specifications [8]-[10] include the following:

Multiple Access Schemes: DL: OFDMA with CP UL: SCFDMA with CP Modulation: UL/DL: QPSK, 16QAM, 64QAM Coding: Convolution code, Rel-6 Turbo code.

21

1.6 Multi Carrier Modulations:

Unlike single carrier systems, OFDM communication systems do not rely on increased symbol rates in order to achieve higher data rates. OFDM is a multicarrier digital modulation scheme. OFDM systems break the available bandwidth into many narrower sub-carriers and transmit the data in parallel streams. Each subcarrier is modulated using varying levels of QAM modulation, e.g. QPSK, QAM, 64QAM or possibly higher orders depending on signal quality. Each OFDM symbol is therefore a linear combination of the instantaneous signals on each of the sub-carriers in the channel .This scheme facilitates efficient use of bandwidth and reduced Inter Symbol Interference (ISI). But another problem is high Peak to Average Power Ratio (PAPR) OFDM symbols .To counter this we use a modified scheme called Single Carrier FDMA (SC-FDMA).The advantages are reduced PAPR and frequency domain equalization [6].

1.7 Objective of Thesis:

The main objectives of thesis are: (1) A comparative study of SCFDMA and OFDMA which are used for uplink and downlink communication in 3GPP LTE system. (2) Comprising of PAPR analysis for both the techniques under different conditions or parameters. (3) Reducing the PAPR of OFDM signal using the clipping and filtering method.

1.8 Scope of Thesis:

The thesis is organized as follows: Chapter 2 presents the overall architectural overview of the LTE system. Chapter 3 discusses about the physical layer in the LTE uplink and its multiple access SCFDMA. Chapter 4 discusses about the physical layer in the LTE downlink and its multiple access OFDMA. Chapter 5 discusses about the basics of PAPR analysis and comparative study of PAPR for SCFDMA & OFDMA and PAPR reduction techniques for OFDM signal. 22

Chapter 6 provides the characteristics of mobile radio channels and different ways to model channel impulse responses. Chapter 7 investigates different channel estimation techniques, they are LS estimator, LMMSE estimator and modified MMSE estimator. Chapter 8 deals with simulations and results under different parametric conditions. Chapter 9 concludes on the entire discussion.

23

Chapter 2 LTE Network Architecture

2.1 Introduction:
In contrast to the circuit-switched model of previous cellular systems, Long Term Evolution (LTE) has been designed to support only packet-switched services. It aims to provide seamless Internet Protocol (IP) connectivity between user equipment (UE) and the packet data network (PDN), without any disruption to the end users applications during mobility. While the term LTE encompasses the evolution of the Universal Mobile

Telecommunications System (UMTS) radio access through the Evolved UTRAN (EUTRAN), it is accompanied by an evolution of the non-radio aspects under the term System Architecture Evolution (SAE), which includes the Evolved Packet Core (EPC) network. Together LTE and SAE comprise the Evolved Packet System (EPS). EPS uses the concept of EPS bearers to route IP traffic from a gateway in the PDN to the UE. A bearer is an IP packet flow with a defined quality of service (QoS) between the gateway and the UE. The EUTRAN and EPC together set up and release bearers as required by applications. This paper provides a comprehensive tutorial of the overall EPS network architecture, giving an overview of the functions provided by the core network (CN) and E-UTRAN. The protocol stack across the different interfaces is explained, along with an overview of the functions provided by the different protocol layers. The end-to-end bearer path along with QoS aspects are also discussed, including a typical procedure for establishing a bearer. The remainder of this paper presents the network interfaces in detail, with particular focus on the E-UTRAN interfaces and the procedures used across these interfaces, including those for the support of user mobility [5].

2.2 Overall Architectural Overview:

EPS provides the user with IP connectivity to a PDN for accessing the Internet, as well as for running services such as Voice over IP (VoIP). An EPS bearer is typically associated with a QoS. Multiple bearers can be established for a user in order to provide different QoS streams or connectivity to different PDNs. For example, a user might be engaged in a voice (VoIP) 24

call while at the same time performing web browsing or FTP download. A VoIP bearer would provide the necessary QoS for the voice call, while a best-effort bearer would be suitable for the web browsing or FTP session. The network must also provide sufficient security and privacy for the user and protection for the network against fraudulent use. This is achieved by means of several EPS network elements that have different roles. Fig 2.1 shows the overall network architecture, including the network elements and the standardized interfaces. At a high level, the network is comprised of the CN (EPC) and the access network E-UTRAN. While the CN consists of many logical nodes, the access network is made up of essentially just one node, the evolved NodeB (eNodeB), which connects to the UEs. Each of these network elements is interconnected by means of interfaces that are standardized in order to allow multi-vendor interoperability. This gives network operators the possibility to source different network elements from different vendors. In fact, network operators may choose in their physical implementations to split or merge these logical network elements depending on commercial considerations. The functional split between the EPC and E-UTRAN is shown in Fig 2.2. The EPC and E-UTRAN network elements are described in more detail below [5].

HSS UE S6a MME

LTE - Uu

PCRF

Rx

S1-MME

S11

Gx S5/S8

Operators IP services ( For ex- IMS, PSS)

SGi

eNodeB

S1-U

S - GW

P - GW

Fig 2.1 The EPS network elements.

25

2.2.1 The core network: The CN (called EPC in SAE) is responsible for the overall control of the UE and establishment of the bearers. The main logical nodes of the EPC are: PDN Gateway (P-GW); Serving Gateway (S-GW); Mobility Management Entity (MME).

In addition to these nodes, EPC also includes other logical nodes and functions such as the Home Subscriber Server (HSS) and the Policy Control and Charging Rules Function (PCRF). Since the EPS only provides a bearer path of a certain QoS, control of multimedia applications such as VoIP is provided by the IP Multimedia Subsystem (IMS) which is considered to be outside the EPS itself. The logical CN nodes are shown in Figure 2.1 and discussed in more detail [5] in the following. PCRF: It is responsible for policy control decision-making, as well as for controlling the ow-based charging functionalities in the Policy Control Enforcement Function (PCEF) which resides in the P-GW. The PCRF provides the QoS authorization (QoS class identier and bitrates) that decides how a certain data ow will be treated in the PCEF and ensures that this is in accordance with the users subscription prole. Home Location Register (HLR): The HLR contains users SAE subscription data such as the EPS-subscribed QoS profile and any access restrictions for roaming. It also holds information about the PDNs to which the user can connect.

This could be in the form of an Access Point Name (APN) (which is a label according to DNS1 naming conventions describing the access point to the PDN), or a PDN Address (indicating subscribed IP address(es). In addition the HLR holds dynamic information such as the identity of the MME to which the user is currently attached or registered. The HLR may also integrate the Authentication Centre (AuC) which generates the vectors for authentication and security keys.

26

eNodeB Inter Cell RRM RB Control Connection Mobility Control Radio Admission Control eNodeB Measurement Configuration & Provision Dynamic Resource Allocation (Scheduler) RRC PDCP RLC MAC S1 PHY S - GW Mobility Anchoring P - GW UE IP Address Allocation Packet Filtering

MME NAS Security Idle State Mobility Handling EPS Bearer Control

Internet E-UTRAN EPC

Fig 2.2: Functional split between E-UTRAN and EPC.

P-GW: The P-GW is responsible for IP address allocation for the UE, as well as QoS enforcement and ow-based charging according to rules from the PCRF. The P-GW is responsible for the ltering of downlink user IP packets into the different QoS based bearers. This is performed based on Traffic Flow Templates (TFTs). The P-GW performs QoS enforcement for Guaranteed Bit Rate (GBR) bearers. It also serves as the mobility anchor for inter-working with non-3GPP technologies such as CDMA2000 and WiMAX networks. 27

 S-GW: All user IP packets are transferred through the S-GW, which serves as the local mobility anchor for the data bearers when the UE moves between eNodeBs. It also retains the information about the bearers when the UE is in idle state (known as ECM- IDLE) and temporarily buffers downlink data while the MME initiates paging of the UE to re-establish the bearers. In addition, the S-GW performs some administrative functions in the visited network such as collecting information for charging (e.g. the volume of data sent to or received from the user), and legal interception. It also serves as the mobility anchor for interworking with other 3GPP technologies such as GPRS and UMTS. MME: The MME is the control node which processes the signaling between the UE and the CN. The protocols running between the UE and the CN are known as the Non-Access Stratum (NAS) protocols. The main functions supported by the MME are classified as: Functions related to bearer management: This includes the establishment, maintenance and release of the bearers, and is handled by the session management layer in the NAS protocol. Functions related to connection management: This includes the establishment of the connection and security between the network and UE, and is handled by the connection or mobility management layer in the NAS protocol layer. NAS control procedures are discussed in more detail in the following section.

2.2.1.1 Non-Access Stratum (NAS) Procedures: The NAS procedures, especially the connection management procedures, are fundamentally similar to UMTS. The main change from UMTS is that EPS allows concatenation of some procedures to allow faster establishment of the connection and the bearers. The MME creates a UE context when a UE is turned on and attaches to the network. It assigns a unique short temporary identity termed the SAE-Temporary Mobile Subscriber Identity (S-TMSI) to the UE which identifies the UE context in the MME. This UE context holds user subscription information downloaded from the HSS. The local storage of subscription data in the MME allows faster execution of procedures such as bearer establishment since it removes the need to consult the HSS every time. In addition, the UE context also holds dynamic information such as the list of bearers that are established and the terminal capabilities.

28

To reduce the overhead in the E-UTRAN and processing in the UE, all UE-related information in the access network can be released during long periods of data inactivity. This state is called EPS Connection Management IDLE (ECM-IDLE). The MME retains the UE context and the information about the established bearers during these idle periods. To allow the network to contact an ECM-IDLE UE, the UE updates the network as to its new location whenever it moves out of its current Tracking Area (TA); this procedure is called a Tracking Area Update. The MME is responsible for keeping track of the user location while the UE is in ECM-IDLE. When there is a need to deliver downlink data to an ECM-IDLE UE, the MME sends a paging message to all the eNodeBs in its current TA, and the eNodeBs page the UE over the radio interface. On receipt of a paging message, the UE performs a service request procedure which results in moving the UE to ECM-CONNECTED state. UE-related information is thereby created in the E-UTRAN, and the bearers are re-established. The MME is responsible for the re-establishment of the radio bearers and updating the UE context in the eNodeB. This transition between the UE states is called an idle-to-active transition. To speed up the idle-toactive transition and bearer establishment, EPS supports concatenation of the NAS and AS procedures for bearer activation. Some inter-relationship between the NAS and AS protocols is intentionally used to allow procedures to run simultaneously rather than sequentially, as in UMTS. For example, the bearer establishment procedure can be executed by the network without waiting for the completion of the security procedure. Security functions are the responsibility of the MME for both signaling and user data. When a UE attaches with the network, a mutual authentication of the UE and the network is performed between the UE and the MME/HSS. This authentication function also establishes the security keys which are used for encryption of the bearers [5].

2.2.2 The access network:

The Access Network of LTE, E-UTRAN, simply consists of a network of eNodeBs, as illustrated in Fig 2.3. For normal user traffic (as opposed to broadcast), there is no centralized controller in E-UTRAN; hence the E-UTRAN architecture is said to be at. The eNodeBs are normally inter-connected with each other by means of an interface known as X2, and to the EPC by means of the S1 interface more specically, to the MME by means of the S1-MME interface and to the S-GW by means of the S1-U interface. The protocols which run between

29

the eNodeBs and the UE are known as the Access Stratum (AS) protocols. The E-UTRAN is responsible for all radio-related functions, which can be summarized briey [5] as: Radio Resource Management: This covers all functions related to the radio bearers, such as radio bearer control, radio admission control, radio mobility control, scheduling and dynamic allocation of resources to UEs in both uplink and downlink. Header Compression: This helps to ensure efficient use of the radio interface by compressing the IP packet headers which could otherwise represent a significant overhead, especially for small packets such as VoIP. Security: All data sent over the radio interface is encrypted. Connectivity to the EPC: This consists of the signaling towards the MME and the bearer path towards the S-GW.

MME / S-GW

MME / S-GW

S1

S1

S1

S1

eNodeB#1

X2

eNodeB#3 E- UTRAN

X2

X2

eNodeB#2

Fig 2.3: Overall E-UTRAN architecture.

30

On the network side, all of these functions reside in the eNodeBs, each of which can be responsible for managing multiple cells. Unlike some of the previous second- and thirdgeneration technologies, LTE integrates the radio controller function into the eNodeB. This allows tight interaction between the different protocol layers of the radio access network, thus reducing latency and improving efficiency. Such distributed control eliminates the need for a high-availability, processing-intensive controller, which in turn has the potential to reduce costs and avoid single points of failure. Furthermore, as LTE does not support soft handover there is no need for a centralized data-combining function in the network.

One consequence of the lack of a centralized controller node is that, as the UE moves, the network must transfer all information related to a UE, i.e. the UE context, together with any buffered data, from one eNodeB to another. Mechanisms are therefore needed to avoid data loss during handover. An important feature of the S1 interface linking the Access Network to the CN is known as S1-ex. This is a concept whereby multiple CN nodes (MME/S-GWs) can serve a common geographical area, being connected by a mesh network to the set of eNodeBs in that area. An eNodeB may thus be served by multiple MME/S-GWs, as is the case for eNodeB#2 in Figure 2.3. The set of MME/S-GW nodes which serves a common area is called an MME/S-GW pool, and the area covered by such a pool of MME/S-GWs is called a pool area. This concept allows UEs in the cell(s) controlled by one eNodeB to be shared between multiple CN nodes, thereby providing a possibility for load sharing and also eliminating single points of failure for the CN nodes. The UE context normally remains with the same MME as long as the UE is located within the pool area.

2.3 Roaming Architecture:

A network run by one operator in one country is known as a Public Land Mobile Network (PLMN). Roaming, where users are allowed to connect to PLMNs other than those to which they are directly subscribed is a powerful feature for mobile networks, and LTE/SAE is no exception. A roaming user is connected to the E-UTRAN, MME and S-GW of the visited LTE network. However, LTE/SAE allows the P-GW of either the visited or the home network to be used, as shown in Fig 2.4. Using the home networks P-GW allows the user to

31

access the home operators services even while in a visited network. A P-GW in the visited network allows a local breakout to the Internet in the visited network [5].

PCRF Rx Gx
HSS

PDN Gateway HPMN VPLMN

MME

SGi

Operators IP Services (e.g. IMS, PSS)

S8

S1-MME

S11

UE

LTE-Uu

E - UTRAN

S1-U

Serving Gateway

Fig 2.4: Roaming architecture for 3GPP accesses with P-GW in home network.

2.4 Inter-Working with Other Networks:

EPS also supports inter-working and mobility (handover) with networks using other Radio Access Technologies (RATs), notably GSM, UMTS, CDMA2000 and WiMAX. The architecture for inter-working with 2G and 3G GPRS/UMTS networks is shown in Fig 2.5. The S-GW acts as the mobility anchor for inter-working with other 3GPP technologies such as GSM and UMTS, while the P-GW serves as an anchor allowing seamless mobility to non3GPP networks such as CDMA2000 or WiMAX. The P-GW may also support a Proxy 32

Mobile Internet Protocol (PMIP) based interface. More details of the radio interface procedures for inter-working are specied [5].

UTRAN

3G-SGSN

S3 MME S1-MME S11 PDN Gateway

S4

UE

LTE-Uu

E - UTRAN

S1-U

Serving Gateway

S5/S8

Fig 2.5: Architecture for 3G UMTS interworking.

2.5 Inter-Radio Access Technologies (RAT) Mobility:

One key element of the design of the first release of LTE is the need to co-exist with other technologies. For mobility from LTE towards UMTS, the handover process can reuse the S1handover procedures described above, with the exception of the STATUS TRANSFER message which is not needed at steps 10 and 11 since no PDCP context is continued. For mobility towards CDMA2000, dedicated uplink and downlink procedures have been introduced in LTE. They essentially aim at tunneling the CDMA2000 signaling between the UE and the CDMA2000 system over the S1 interface, without being interpreted by the eNodeB on the way. The UPLINK S1 CDMA2000 TUNNELLING message presented in Fig 2.6 also includes the RAT type in order to identify which CDMA2000 RAT the tunneled CDMA2000 message is associated with in order for the message to be routed to the correct node within the CDMA2000 system.

33

eNodeB

MME

UPLINK S1 CDMA2000 TUNNELING

Fig 2.6: Uplink S1 CDMA2000 tunneling procedure.

2.6 Connected Mode Inter-RAT Mobility:

The overall procedure for the control of mobility is explained in this section;

2.6.1 Handover to LTE: The procedure for handover to LTE is largely the same as the procedure for handover within LTE, so it is not necessary to repeat the details here. The main difference is that upon handover to LTE the entire AS-configuration needs to be signaled, whereas within LTE it is possible to use delta signaling, whereby only the changes to the configuration are signaled. If ciphering had not yet been activated in the previous RAT, the E-UTRAN activates ciphering, possibly using the NULL algorithm, as part of the handover procedure. The EUTRAN also establishes SRB1, SRB2 and one or more DRBs (i.e. at least the DRB associated with the default EPS bearer).

2.6.2 Mobility from LTE: The procedure for mobility from LTE to another RAT supports both handover and Cell Change Order (CCO), possibly with Network Assistance (NACC Network Assisted Cell Change). The CCO/NACC procedure is applicable only for mobility to GERAN. Mobility

34

from LTE is performed only after security has been activated. The procedure is illustrated in Fig 2.7.

UE

Source eNodeB

Target RAN

Measurement Report

Handover Preparation Mobility From EUTRA Command

Handover Complete OR Connection Establishment Fig 2.7: Mobility from LTE.

1. The UE may send a Measurement Report message.

2. In case of handover (as opposed to CCO), the source eNodeB requests the target RAN node to prepare for the handover. As part of the handover preparation request the source eNodeB provides information about the applicable inter-RAT UE capabilities as well as information about the currently-established bearers. In response, the target RAN generates the handover command and returns this to the source eNodeB.

3. The source eNodeB sends a Mobility From EUTRA Command message to the UE, which includes either the inter-RAT message received from the target (in case of handover), or the target cell/frequency and a few inter-RAT parameters (in case of CCO).

4. Upon receiving the Mobility From EUTRA Command message, the UE starts the timer T304 and connects to the target node, either by using the received radio configuration (handover) or by initiating connection establishment (CCO) in accordance with the applicable specifications of the target RAT. 35

Chapter 3 Physical Layer in the LTE Uplink

3.1 Introduction:
SC-FDMA combines the desirable characteristics of OFDM with the low PAPR of singlecarrier transmission schemes. Like OFDM, SC-FDMA divides the transmission bandwidth into multiple parallel subcarriers, with the orthogonality between the subcarriers being maintained in frequency-selective channels by the use of a Cyclic Prex (CP) or guard period. The use of a CP prevents Inter-Symbol Interference (ISI) between SC-FDMA information blocks. It transforms the linear convolution of the multipath channel into a circular convolution, enabling the receiver to equalize the channel simply by scaling each subcarrier by a complex gain factor. However, unlike OFDM, where the data symbols directly modulate each subcarrier independently (such that the amplitude of each subcarrier at a given time instant is set by the constellation points of the digital modulation scheme), in SC-FDMA the signal modulated onto a given subcarrier is a linear combination of all the data symbols transmitted at the same time instant. Thus in each symbol period, all the transmitted subcarriers of an SC-FDMA signal carry a component of each modulated data symbol. This gives SC-FDMA its crucial single-carrier property, which results in the PAPR being signicantly lower than pure multicarrier transmission schemes such as OFDM [5].

3.2 LTE Uplink Requirements:

While many of the requirements for the design of the LTE uplink physical layer and multipleaccess scheme are similar to those of the downlink, the uplink also poses some unique challenges. Some of the desirable attributes for the LTE uplink include [5]: Orthogonal uplink transmission by different User Equipment (UEs), to minimize intracellular interference and maximize capacity. Flexibility to support a wide range of data rates, and to enable data rate to be adapted to the SINR (Signal-to-Interference plus Noise Ratio). Sufficiently low Peak-to-Average Power Ratio (PAPR) of the transmitted waveform, to avoid excessive cost, size and power consumption of the UE Power Amplier (PA). 36

 Ability to exploit the frequency diversity afforded by the wideband channel (up to 20 MHz), even when transmitting at low data rates. Support for frequency-selective scheduling. Support for advanced multiple-antenna techniques, to exploit spatial diversity and enhance uplink capacity. The multiple-access scheme selected for the LTE uplink so as to full these principle characteristics is Single-Carrier Frequency Division Multiple Access (SC-FDMA). A major advantage of SC-FDMA over the Direct-Sequence Code Division Multiple Access (DSCDMA) scheme used in LTE is that it achieves intra-cell orthogonality and low PAPR.

3.3 SC-FDMA Principles:

3.3.1 SC-FDMA transmission structure: An SC-FDMA signal can, in theory, be generated in either the time-domain or the frequencydomain . Although the two techniques are duals and functionally equivalent, in practice, the time-domain generation is less bandwidth-efficient due to time-domain ltering and associated requirements for lter ramp-up and ramp-down times [5]. Nevertheless, we describe both approaches here to facilitate understanding of the principles of SC-FDMA in both domains.

3.3.2 Time-domain signal generation: Time-domain generation of an SC-FDMA signal is shown in Fig 3.1. It can be seen to be similar to conventional single-carrier transmission. The input bit stream is mapped into a single-carrier stream of QPSK or QAM symbols, which are grouped into symbol-blocks of length M. This may be followed by an optional repetition stage, in which each block is repeated L times, and a user-specic frequency shift, by which each users transmission may be translated to a particular part of the available bandwidth. A CP is then inserted. After ltering (e.g. with a root-raised cosine pulse-shaping lter), the resulting signal is transmitted. Different users transmissions, using different repetition factors or bandwidths, remain orthogonal on the uplink when the following conditions are met [5]: The users occupy different sets of subcarriers. This may in general be accomplished either by introducing a user-specic frequency shift (typically for the case of localized transmissions) or alternatively by arranging for different users to occupy interleaved sets of 37

subcarriers (typically for the case of distributed transmissions). The latter method is known in the literature as Interleaved Frequency Division Multiple Access (IFDMA).

S/P Converter

DS-Spreading (Optional)

Bit to Constellation Mapping

Userspecific frequency shift

Fig 3.1: SC-FDMA time-domain transmit processing.

The received signals are properly synchronized in time and frequency. The CP is longer than the sum of the delay spread of the channel and any residual timing synchronization error between the users. The SC-FDMA time-domain generated signal has a similar level of CM/PAPR as pulse-shaped single-carrier modulation. ISI in multipath channels is prevented by the CP, which enables efficient equalization at the receiver by means of a Frequency Domain Equalizer (FDE).

3.4 SC-FDMA Frame Structure:

The generic frame structure for the SC-FDMA uplink is shown [11] in Fig 3.2.
One Radio Frame, Tf = 10ms One Slot Tslot = 0.5 ms

#0

#1

#2

#3

# 18

One Subframe

Fig 3.2: Generic frame structure (TDD or FDD)

38

Pulse-shape filter

# 19

Transmission circuitry

Incoming bit stream

User-specific Block repetition

Add CP

The generic slot structure with a normal cyclic prefix is shown in Fig 3.3. A slot with an extended cyclic prefix contains only 6 long blocks.
One Slot = 0.5 ms
CP LB0 CP LB0 CP LB0 CP LB0 CP LB0 CP LB0 CP LB0

Fig 3.3: Slot structure

- CP = Cyclic prefix (guard interval) - LB = Long block (for data symbol)

3.5 Uplink SC-FDMA Parameters:

Table 3.1: Uplink parameters for SC-FDMA transmission [11]. 20 MHz

Transmission BW Slot duration (generic frame structure) Slot duration (alternative frame structure) CP duration ms / no. of subcarriers (generic frame structure) CP duration ms / no. of subcarriers (alternative frame structure) ms Occupied subcarriers FFT size ms Occupied subcarriers FFT size

1.25 MHz

2.5 MHz

5 MHz
0.5 ms

10 MHz

15 MHz

0.675 ms 3.65/7 or 7.81/15 6.25/12 or 10.4/20 3.91/15 or 5.99/23 6.51/25 or 8.58/33 4.04/31 or 5.08/39 6.64/51 or 7.67/59 66.67 4.1/63 or 4.62/71 6.71/103 or 7.22/111 4.12/95 or 4.47/103 6.77/156 or 7.11/164 4.13/127 or 4.39/135 6.71/206 or 6.97/214

Long block (LB) size

75 128

150 256

300 512
33.33

600 1024

900 1536

1200 2048

Short block (SB) size

38 64

75 128

150 256

300 512

450 768

600 1024

39

3.6 Modulation:
There are no harmonization problems between the downlink and the uplink in terms of frame structure and modulation parameters. The modulation scheme that is used can be QPSK, 16QAM or 64QAM according to the channel quality. Specifically, the uplink symbols enter a serial/ parallel converter and then into a FFT block. The result is mapped onto the available sub-carriers. Later, a N point IFFT is applied, the cyclic prefix is added and, finally, this result enters a parallel to serial converter [12].

3.7 Implementation of the SC-FDMA Transceiver:

Different transmitters (users) are assigned different Fourier coefficients. This assignment is carried out in the mapping and demapping blocks. The transmitter of LTE uplink is designed as illustrated in fig 3.4. The receiver side includes one demapping block, one IDFT block and one detection block for each user signal to be received. The receiver of LTE uplink is designed as illustrated in fig 3.5 [13]. SC-FDMA is a new multiple access technique that utilizes single carrier modulation, DFT-spread orthogonal frequency multiplexing, and frequency domain equalization. It has a similar structure and performance as OFDM. SCFDMA is currently adopted as the uplink multiple access scheme for 3GPP LTE. Transmitter and receiver structure for SC-FDMA are given in Figures 3.4 and 3.5. It is evident from the figures that SC-FDMA transceiver has similar structure as a typical OFDM system except the addition of a new DFT block before subcarrier mapping. Hence, SC-FDMA can be considered as an OFDM system with a DFT mapper.

S/P Conversion

N-FFT

Sub-carrier mapping

M-IFFT

P/S Conversion

Add CP

Fig 3.4: Block diagram of the SC-FDMA transmitter in LTE.

P/S Conversion

N-IFFT

Sub-carrier demapping

M-FFT

S/P Conversi on

Remove CP

Fig 3.5: Block diagram of the SC-FDMA receiver in LTE. 40

3.8 LTE Uplink Physical Channels:

Physical Uplink Control Channel (PUCCH): It provides control signaling information such as ACK/NACK information, CQI (channel quality indication) reports, RI (rank indication) and other formats. Physical Uplink Shared Channel (PUSCH): It is the Uplink counterpart of PDSCH. Physical Random Access Channel (PRACH): It is used for random access functions. Through this, the downlink and uplink propagation delays are not known. As a result, the transmission cannot get synchronized [12].

3.9 LTE Uplink Transport Channels:

Uplink Shared Channel (UL-SCH) : It is the most important channel for uplink data transfer used by several logical channels. Random Access Channel (RACH) : It is used for random access requirements [12].

41

Chapter 4 Physical Layer in the LTE Downlink

4.1 Introduction:
One of the main changes in the LTE system compared to 3G-UMTS is the physical layer. In third generation systems, Wideband Code Division Multiple Access (WCDMA) is the most widely adopted technology. A highlight of the characteristics of the UMTS before Release 7 is listed below [14]: - User information bits are spread over a wide bandwidth by multiplying the user data with a spreading code. The use of variable spreading factor allows a variation of the bit rate. - The bandwidth is 5 MHz. The chip rate used is 3.84 Mbps. A network operator can deploy multiple 5 MHz bands to increase capacity. - The frame length is 10 ms. During this phase, the user data rate is kept constant. However, the data rate among the users can change from frame to frame. In the LTE system, this will be very different. The new system will present an OFDM based structure. The main aspects important for channel estimation in the physical layer are presented in the following section. In the LTE only packet-switched transmission is utilized. OFDMA fits perfectly into packet-switched transmission, since different number of subcarriers (RBs) can be assigned to different users, in order to support differentiated Quality of Service (QoS). The scheduling is dynamic and performed for each sub-frame, hence the number of RBs can be adjusted dynamically depending on the channel quality.

4.2 LTE Downlink Requirements:

The technique of OFDM is based on the technique of frequency division multiplexing (FDM). The OFDM technique differs from traditional FDM by having subcarriers, which are orthogonal to each other. The modulation technique used in an OFDM system helps to overcome the effects of a frequency selective channel. A frequency selective channel occurs when the transmitted signal experiences a multipath environment. Under such conditions, a given received symbol can be potentially corrupted by a number of previous symbols. This

42

effect is commonly known as inter-symbol interference (ISI). To avoid such interference, the symbol duration has to be much larger than the delays caused by multipath channel. Hence each symbol is prolonged with a copy of its tail denoted as cyclic prefix (CP) such that the ISI is minimized. Also, the spectral efficiency of the OFDM modulation technique is superior to FDM since the subcarriers are overlapping, but orthogonal. The frequency spacing between the Subcarriers =

is either 15 kHz or 7.5 kHz according to working

assumption in Release 8 [5]. In contrast to an OFDM transmission scheme, OFDMA allows multiple users to share the available bandwidth. Each user is assigned a specic timefrequency resource referred as resource block (RB). The fundamental principle of the Evolved UMTS Terrestrial Radio Access (E-UTRA) is that the data channels are shared channels, i.e. for each transmission time interval (TTI) of 1ms, a new scheduling decision is made at eNodeB regarding which users are assigned to which time/frequency resources during this transmission time interval [14].

4.3 OFDM Principles:

4.3.1 Orthogonal multiplexing principle: A high-rate data stream typically faces a problem in having a symbol period Ts much smaller than the channel delay spread Td if it is transmitted serially. This generates Inter- symbol Interference (ISI) which can only be undone by means of a complex equalization procedure. In general, the equalization complexity grows with the square of the channel impulse response length. In OFDM, the high-rate stream of data symbols is rst serial-to-parallel converted for modulation onto M parallel subcarriers as shown in Fig 4.1. This increases the symbol duration on each subcarrier by a factor of approximately M, such that it becomes significantly longer than the channel delay spread. This operation has the important advantage of requiring a much less complex equalization procedure in the receiver, under the assumption that the time-varying channel impulse response remains substantially constant during the transmission of each modulated OFDM symbol. This operation has the important advantage of requiring a much less complex equalization procedure in the receiver, under the assumption that the time-varying channel impulse response remains substantially constant during the transmission of each modulated OFDM symbol. Figure 5.3 shows how the resulting long symbol duration is virtually

43

unaffected by ISI compared to the short symbol duration, which is highly corrupted. Figure 5.4 shows the typical block diagram of an OFDM system [5].

Low Symbol Rate

S/P

Fig 4.1: Serial-to-parallel conversion operations for OFDM.

xk[N-G] Cyclic Prefix xk[N-1] Sk[0] Xk[0] xk[0] Sk[1] S/P Sk[N-2] Sk[N-1] Xk[N-2] Xk[N-1] Xk[1] IFFT xk[N-G] xk[N-1] xk[1] P/S DAC

Fig 4.2: OFDM Transmitter The signal to be transmitted is dened in the frequency domain. A Serial to Parallel (S/P) converter collects serial data symbols into a data block S k = [Sk [0] ,Sk [1] ,...,Sk [M 1]]T of dimension M, where the subscript k is the index of an OFDM symbol (spanning the M subcarriers). The M parallel data streams are rst independently modulated resulting in the complex vector Xk = [Xk [0] ,Xk [1] , ..., Xk [M 1]]T . Note that in principle it is possible to 44

use different modulations (e.g. QPSK or 16QAM) on each sub-carrier; due to channel frequency selectivity, the channel gain may differ between sub-carriers, and thus some subcarriers can carry higher data-rates than others.
rkCP[0] Cyclic Prefix removal rkCP[G-1] rkCP[G] = rk[0] ADC rkCP[G+1] = rk[1] Yk[0] Yk[1]

S/P

rkCP[N+G-2] = rk[N-2]

FFT Yk[N-2]

rkCP[N+G-1] = rk[N-1]

Yk[N-1]

Fig 4.3: OFDM receiver

The vector of data symbols Xk then passes through an Inverse FFT (IFFT) resulting in a set of N complex time domain samples xk = [xk[0],...,xk[N 1]]T . In a practical OFDM system, the number of processed sub- carriers is greater than the number of modulated sub-carriers (i.e. N M), with the un-modulated sub-carriers being padded with zeros.

TCP

Tu

TCP

Tu

Fig 4.4: OFDM cyclic prex insertion. The next key operation in the generation of an OFDM signal is the creation of a guard period at the beginning of each OFDM symbol, to eliminate the remaining impact of ISI caused by 45

multipath propagation. The guard period is obtained by adding a Cyclic Prex (CP) at the beginning of the symbol xk. The CP is generated by duplicating the last G samples of the IFFT output and appending them at the beginning of xk. This yields the time domain OFDM symbol [xk[N G], ..., xk[N 1], xk[0], ...,xk[N 1]] T ,as shown in Fig 4.3. To avoid ISI completely, the CP length G must be chosen to be longer than the longest channel impulse response to be supported. The CP converts the linear (i.e. a-periodic) convolution of the channel into a circular (i.e. periodic) one which is suitable for DFT processing. This important feature of CP used in OFDM is explained more formally later in this section. The output of the IFFT is then Parallel-to-Serial (P/S) converted for transmission through the frequency-selective channel. At the receiver, the reverse operations are performed to demodulate the OFDM signal. Assuming that time- and frequency-synchronization is achieved, a number of samples corresponding to the length of the CP are removed, such that only an ISI-free block of samples is passed to the DFT. If the number of subcarriers N is designed to be a power of 2, a highly efficient FFT implementation may be used to transform the signal back to the frequency domain. Among the N parallel streams output from the FFT, the modulated subset of M subcarriers are selected and further processed by the receiver. Let x(t) be the signal symbol transmitted at time instant t . The received signal in a multipath environment is then given by r(t) = x(t) h(t) + z(t) (4.1)

where h(t) is the continuous-time impulse response of the channel, represents the convolution operation and z(t) is the additive noise. Assuming that x(t) is band-limited to [ 1 2 , 1 2 ], the continuous-time signal x(t) can be sampled at sampling rate Ts such that the Nyquist criterion is satised. As a result of the multipath propagation, several replicas of the transmitted signals arrive at the receiver at different delays [5].

4.3.2 Importance of orthogonality: The orthogonal part of OFDM name indicates there is some mathematical relationship between frequencies in sub bands. Introduction of guard bands reduces the spectral

efficiency. So to enhance this efficiency, the carriers in OFDM signals are arranged in a manner such that individual carriers overlap and the signals can still be received without carrier interference. Mathematically, two signals are orthogonal if 46

if p = q

(4.2)

if p q

Where * denotes the complex conjugate and interval [a b] is a symbol period [16]. An OFDM signal consists of a sum of subcarriers that are modulated by using BPSK, QPSK or QAM. Mathematically, each carrier can be described as a complex wave: = { + ()} (4.3)

OFDM being carrying many carriers, its signal representation is: = Where, = + This is a continuous signal. If we consider the waveforms of each component of the signal over one symbol period, then Ac(t) and fc(t) take on fixed values, which depends on the frequency of that particular carrier, and so can be rewritten as: = and =
1 =1 =0

{ + ()}

(4.4)

if now the signal is sampled at T time period, then the resulting signal becomes: =
1 =1 { + } =0

(4.5)

At this point, we restricted the time of analysis upto N samples. But its convenient to sample over one data symbol period. Thus we have: = NT If we simplify eqn. 4.5, without the loss of generality by letting o = 0, then the signal becomes: =
1 =1 =0

( )

(4.6)

This can now be compared with the general form of inverse Fourier Transform: =
1 =1 ( ) 2 / =0

(4.7)

Eqns 3.5 and 3.6 are equivalent if: 47

= =

(4.8)

This is the same condition that was required for orthogonality. Thus, maintaining orthogonality is that the OFDM signal can be defined by using Fourier transform procedures [16].

4.3.3 Guard interval: Individual sub channels can be completely separated by the FFT at the receiver when there are no ISI and ICI introduced by channel distortion. Practically these conditions cannot be obtained. Since the spectra of an OFDM signal is not strictly band limited, linear distortion such as multipath fading cause sub channel to spread energy in the adjacent channels [16]. This problem can be solved by increasing symbol duration. One way to prevent ISI is to create a cyclically extended guard interval, where each symbol is preceded by a periodic extension of the signal itself. The total symbol duration being increased to T Total = Tg + T. When Tg is longer than the channel impulse response, the ISI can be eliminated. Since the insertion of guard interval will reduce data throughput, T g is usually less than T/4. The main reasons to use a cyclic prefix for the guard band interval are [16]: 1. To maintain the receiver carrier synchronization. 2. Cyclic convolution can still be applied between the OFDM signal and the channel response to model the transmission systems.

CP

Symbol

CP

Symbol

Fig 4.5: Insertion of cyclic prefix

4.4 OFDM Frame Structure:

The structure of the radio frame, illustrated in Fig 4.6, is described in the current study from 3GPP. It should be noticed that for time division duplex (TDD), sub-frames for uplink and downlink purpose should be assigned. Other frame structures are proposed in order to make the structure compatible with the present structure used in 3G. For simplicity it is chosen to work with the illustrated generic frame structure. The duration of one frame is 10 ms and is 48

composed of 20 slots of 0.5 ms, where one sub-frame consists of two slots. The number of OFDM symbols in one slot Nsym depends on the chosen length of the cyclic prefix (CP) and can be either 6 (long CP) or 7 (short CP).
Tframe = 10ms

20 slots

Tslot = 0.5ms Tsubframe = 1ms 6 or 7 OFDM symbols

Fig 4.6: OFDM Frame structure in LTE [14]. A radio frame is divided into 20 slots of 0.5 ms each having 6 or 7 OFDM symbols. Two slots make one sub frame, which corresponds to the minimum downlink TTI.

4.5 Downlink OFDM Parameters:

The parameters used for downlink are listed in Table 4.1. The subcarrier frequency spacing =

= 15 kHz is used, and it is always constant, hence fs and NIFFT are proportional.

The downlink parameters for fspace = 7.5 kHz are not yet defined [14]. The number of OFDM symbols Nsym per slot depends on the length of the CP as described in section 2.2. If 128point IFFT and short CP is used, the first 6 OFDM symbols have a CP of 9 samples and the last symbol a CP of 10 samples, such that the duration of the sub-frame of 0.5ms is preserved. Not all subcarriers are occupied, in Release 7 [15] approximately 2/3 of the total frequency band is used. According to technical specications in Release 8 [14] the number of used subcarriers (here denoted as NBW) can be varied. The values of NBW however are not specied. In this project the values NBW are the same as in Release 7. Other downlink parameters than number of FFT-points and sampling frequency are not yet determined, but the above assumption is used for evaluation purpose in 3GPP, hence these parameters are also used in the project.

49

Table 4.1: Downlink parameters for OFDM transmission.

Transmission BW Subframe duration Tsub Sub-carrier spacing fspace Sampling frequency fs FFT size NIFFT Number of occupied subcarriers NBW Number of OFDM symbols per subframe (short/long CP) 1.25 MHz 2.5 MHz 5 MHz 10 MHz 0.5 ms 15 KHz 15.36MHz 1024 600 7/6 (4.69/18) 6 (5.21/20) 1 (16.67/64 ) (4.69/36) 6 (5.21/40) 1 (16.67/12 8) (4.69/72) 6 (5.21/80) 1 (16.67/25 6) (4.69/108) 6 (5.21/120) 1 (16.67/38 4) (4.69/144) 6 (5.21/160) 1 (16.67/51 2) 15 MHz 20 MHz

1.92 MHz 128 75

3.84 MHz 256 150

7.68MHz 512 300

23.04 MHz 1536 900

30.72MHz 2048 1200

Short CP length (s / sample) Long

(4.69/9)6 (5.21/10) 1

(16.67/32)

4.6 Mapping of Subcarriers:

The subcarriers are mapped into the frequency spectrum as illustrated in Fig 4.7. According to Table 2.1, NBW is 75/150/300/600/900/1200 when the transmission bandwidth is 1.25/2.5/5/10/15/20 MHz.

Unused Subcarriers

Nn

Nn+1

NBW

Unused Subcarriers

Fig 4.7: Placement of occupied subcarriers [15]. NBW and Nn are the total number of occupied subcarriers and the number of carriers in the negative spectrum respectively.

Since the occupied subcarriers are centered around the frequency 0, half of the occupied subcarriers are placed in the negative spectrum and the other half in the positive spectrum. Let us denote the occupied subcarriers in the negative spectrum as {1, . . . ,Nn} and in the positive spectrum as {Nn + 1, . . . ,NBW}, where Nn is 37/75/150/300/450/600 [14]. The unused carriers are placed at the edges of the spectrum such that the utilized bandwidth is less

50

than the specied bandwidth. This can be based on reducing the requirements for the analog lters at the transmitter and receiver side.

4.7 Implementation of the OFDM Transceiver:

Based on the mentioned information on the physical layer, a structure of the transmitter in LTE is designed as illustrated on Fig 4.8. The transmitter is based on conventional OFDM system structure. The structure of the implemented receiver is depicted in Fig 4.9.

Reference Symbol Insertion

S/P converter

P/S converter

Binary Source Generator

CP insertion

Tx

M-QAM Modulator

IFFT

Signal

Figure 4.8: Block diagram of the OFDM transmitter in LTE.

Reference Symbol Removal

Channel estimation & Equalization

P/S converter

S/P converter

Computation

CP Removal

M-QAM Demodulator

Raw BER

Rx Signal

Fig 4.9: Block diagram of the OFDM receiver in LTE.

4.7.1 Binary Source Generator: The binary source generator generates the signal randomly. The number of the generated binary symbols depends on the modulation scheme, i.e. the number of bits per QAM-symbol and the number of subcarriers [14].

4.7.2 Modulation: During modulation it is necessary to normalize the transmitted symbols in order to adjust the signal-to-noise ratio. The normalization is achieved by scaling the symbols as listed in Table 4.2.

51

FFT

Table 4.2: Normalization factor for M-QAM modulation schemes in E-UTRA downlink [14]. Modulation 4-QAM 16-QAM 64-QAM Knorm 1 2 1 10 1 64

4.8 Downlink Data Transmission:

The transmitted signal in each slot is described by a resource grid of NBW subcarriers and Nsym OFDM symbols. In order to achieve multiple accesses, bandwidth is allocated to the UEs in terms of resource blocks. A physical resource block, NRB consists of 12 consecutive subcarriers in the frequency domain. In the time domain, a physical resource block consists of Nsym consecutive OFDM symbols, Nsym is equal to the number of OFDM symbols in a slot. The resource block size is the same for all bandwidths; hence the number of available physical resource blocks depends on the bandwidth. Depending on the required data rate, each UE can be assigned one or more resource blocks in each transmission time interval of 1 ms. The scheduling decision is done at the NodeB. The user data is carried on the Physical Downlink Shared Channel (PDSCH). Downlink control signaling on the Physical Downlink Control Channel (PDCCH) is used to transport the scheduling decisions to individual UEs. The PDCCH is placed in the rst OFDM symbols of a slot [14]. 4.8.1 Modulation: According to the working assumptions for PDSCH in Release 8, the transmitted bits are modulated using quadrature amplitude modulation (QAM). The available modulation schemes are 4-QAM, 16-QAM, and 64-QAM [15].

4.8.2 Downlink reference signal structure: The downlink reference signal structure is important for cell search and channel estimation. Resource elements in the time-frequency domain are carrying the reference signal sequence, which is predefined for each cell. The reference symbols are placed in the rst OFDM 52

symbol of one slot and on the third last OFDM symbol. The spacing between the reference symbols is always 6 subcarriers [15] and the norm is always 1 no matter which modulation scheme is utilized for the data symbols. In the LTE the eNodeBs and UEs can have 2 or 4 antennas and when two or more transmitter antennas are applied, the reference symbols are transmitted such that they are orthogonal in space.
Subcarriers

X
One Slot Duration TSlot

. 1st OFDM symbol . .

. . . 6th OFDM symbol

Reference symbol Antenna 1

vacant resource element

Subcarriers

X
One Slot Duration TSlot

1st OFDM symbol

6th OFDM symbol

Antenna 2

Fig 4.10: The reference symbol structure for one slot with 6 OFDM symbols using two antennas. Note that only the used subcarriers are depicted. In this thesis we consider one antenna and makes use of the reference symbol structure depicted for antenna 1.

The orthogonality in space is obtained by letting all other antennas be silent in the resource element in which one antenna transmits a reference symbol [14]. Figure 2.4 shows the positions of the reference symbols for transmission with two antennas as an example. When 53

antenna 1 transmits a reference symbol, antenna 2 is silent and vice versa. This thesis considers one antenna and makes use of the reference symbol structure depicted for antenna 1 on Fig 4.10. The reference signal sequence also carries the cell identity. The reference signal sequence is generated as a symbol-by-symbol product of an orthogonal sequence (OS) ROS C3402 (3 different sequences are predefined) and a pseudo-random sequence (PRS) RPRS R3402 (170 different sequences are predefined). Each cell identity corresponds to a unique combination of one orthogonal sequence ROS and one pseudorandom sequence RPRS, allowing 510 different cell identities [14]. Frequency hopping can also be applied to the downlink reference signals. The frequency hopping pattern has a period of one frame duration.

4.8.3 Cell search: During cell search, different types of information need to be identied by the UE such as radio frame timing, frequency, cell identification, overall transmission bandwidth, antenna conguration, cyclic prex length. Besides the reference symbols, synchronization signals are therefore needed during cell search. In E-UTRA (Evolved UMTS Terrestrial Radio Access) the synchronization acquisition and the cell group identifier are obtained from different synchronization channels (SCH). A primary synchronization channel (PSCH) for synchronization acquisition and a secondary synchronization channel (SSCH) for cell group identification have a predened structure. They are transmitted on the 72 subcarriers centered around subcarrier at frequency f = 0 within the same predened slots (1st and 11th slot in one frame). PSCH and SSCH are however placed on the second last and third last OFDM symbol respectively [14]. Hence cell search is always performed using the 72 central subcarriers independent of the overall transmission bandwidth.

4.9 Latency Requirement:

The user plane latency should be below 5 ms. For the downlink case the user plane is dened in terms of a one-way transit time between a packet being available at the IP layer at the NodeB and the availability of this packet at IP layer at the UE. The NodeB provides the interface towards the core network. From channel estimation point of view a latency below 5 ms results in a block length less than 5ms for channel estimation purpose [14].

54

Chapter 5 PAPR Calculation for SCFDMA & OFDMA

5.1 Introduction:
In order to transition from today's 3rd generation (3G) communications systems to meet the needs of 4th generation (4G) systems, the 3rd Generation Partnership Project (3GPP) has released the Long Term Evolution (LTE) specification. Among the numerous differences between these generations are changes in the physical layer, specifically in the modulation and multiple access schemes. While its parent generation relied on variations of Code Division Multiple Access (CDMA), LTE implements Orthogonal Frequency Division Multiplexing (OFDM) for its downlink and Single-Carrier Frequency-Division Multiple Access (SC-FDMA) for its uplink. The purpose of this project is to investigate the reasoning for this discord between uplink and downlink modulation schemes; specifically, why Orthogonal Frequency Division Multiple-Access (OFDMA) was not used as the uplink. OFDMA and SC-FDMA are the multiple-access versions of OFDM and a similar modulation scheme, Single-Carrier Frequency-Domain Equalization (SC-FDE). In order to compare the differences between the multiple-access methods, it is important to first cover the differences between the modulation schemes.

5.2 SCFDMA:
For uplink, SC-FDMA is selected as a basic multiple access scheme for LTE physical layer. SC-FDMA is also a multi-carrier scheme that re-uses many of the functional blocks of OFDMA. The main advantage of SC-FDMA is its low PAPR which is a useful parameter for uplink [15]. OFDMA has small frequency channels, each of which is assigned to a specific symbol. These symbols are transmitted simultaneously as in figure 2.4. As it was mentioned before, prior to transmission over the air all the multiple frequency channels are added together which creates an uncontrollable signal with high peaks. To handle this uncontrollable signal we have to use more power. Using more power is not a problem for downlink however it is one of the main issues in uplink since it increases mobile costs and decreases battery life. Because OFDMA transmits many symbols at a time, we need more power for effective transmission. So as a 55

solution, SC-FDMA decrease the number of symbols transmitted per time, which brings the uncontrollable signal to a manageable levels. Use of wider bandwidth reduces symbol transmission time. For more clarity this can be seen in Fig 5.1 [16].

Bs Hz

OFDMA

Bs Hz

SC-FDMA

T seconds

Fig 5.1: Difference between channel representations between OFDMA and SCFDMA.

3GPP is working on a modified form of OFDMA for uplink transmissions in LTE (long term evolution) of cellular systems. An alternative approach was sought known as Single Carrier Frequency Division Multiple Access (SCFDMA). As in OFDMA, the transmitters in an SCFDMA system use different orthogonal frequencies (subcarriers) to transmit information symbols. However, they transmit the subcarriers sequentially, rather than in parallel. This reduces envelope fluctuation relative to OFDMA. So SCFDMA has inherently low PAPR than OFDMA. But now it has the problem of ISI. It can be removed by adaptive channel equalization algorithms in the frequency domain [18]. Time domain equalization is very complex because of long channel impulse response in time domain and large tap size of filters. But using Discrete Fourier Transform (DFT) in frequency domain its much easier because DFT size doesnt increase linearly with channel response.

56

5.2.1 Block diagram of SC-FDMA: SC-FDMA uses an additional N-point DFT stage at transmitter and an N-point IDFT stage at receiver. The basic block diagram of SC-FDMA transmitter and receiver is shown in Fig 5.2. The input to transmitter is a stream of modulated symbols. In SC-FDMA, the data is mapped into signal constellation according to the QPSK, 16-QAM, or 64-QAM modulation, depending upon the channel conditions similarly as in OFDMA. Whereas, the QPSK/QAM symbols do not directly modulate the subcarriers. These symbols passes through a serial to parallel converter followed by a DFT block that produce discrete frequency domain representation of the QPSK/QAM symbols. Pulse shaping is followed by DFT element, but it is optional and sometimes needs to shape the output signal from DFT. If pulse shaping is active then in the actual signal, bandwidth extension occurs. The discrete Fourier symbols from the output of DFT block are then mapped with the subcarriers in subcarrier mapping block. After mapping this frequency domain modulated subcarriers pass through IDFT for time domain conversion. The rest of transmitter operation is similar as OFDMA.
Carrier
S/P Conversion Subcarrier mapping P/S Conversion

N-FFT

M-IFFT

Add CP

Tx

Channel

P/S Conversion

N-IFFT

Subcarrier demapping

M-FFT

S/P Conversion

Remove CP

Rx Carrier

Fig. 5.2: Tx and Rx structure of SCFDMA (M > N) SC-FDMA receiver is shown in Fig 5.2. It is almost same as conventional OFDMA with additional blocks of subcarrier demapping, IDFT and optional shaping filter. This filter corresponds to the spectral shaping used in the transmitter. The subcarrier demapping of Mmapped subcarrier results N-discrete signals. In the end, IDFT converts the SC-FDMA signal 57

to the signal constellation. In uplink transmission of LTE, there are some additional data carrying signals such as; reference signal, random access preamble and control signal etc. These signals are characterized as sequence signaling and have constant amplitude with zero autocorrelation. In contrast with data carrying signals, these signals are not part of SC-FDMA modulation scheme [19].

5.3 OFDM:
The choice of an appropriate modulation and multiple-access technique for mobile wireless data communications is critical to achieving good system performance. In particular, typical mobile radio channels tend to be dispersive and time-variant, and this has generated interest in multicarrier modulation. In general, multicarrier schemes subdivide the used channel bandwidth into a number of parallel sub-channels as shown in Fig 5.3(a). Ideally the bandwidth of each sub-channel is such that they are each non-frequency-selective (i.e. having a spectrally-at gain); this has the advantage that the receiver can easily compensate for the sub-channel gains individually in the frequency domain.

(a)

Saving in spectrum

(b) Fig 5.3: Spectral efficiency of OFDM compared to classical multicarrier modulation [5]: (a) classical multicarrier system spectrum; (b) OFDM system spectrum. 58

Orthogonal Frequency Division Multiplexing (OFDM) is a special case of multicarrier transmission which is highly attractive for implementation. In OFDM, the non-frequencyselective narrowband sub-channels into which the frequency-selective wideband channel is divided are overlapping but orthogonal, as shown in Figure 5.3(b). This avoids the need to separate the carriers by means of guard-bands, and therefore makes OFDM highly spectrally efficient. The spacing between the sub-channels in OFDM is such they can be perfectly separated at the receiver [5]. This allows for a low-complexity receiver implementation, which makes OFDM attractive for high-rate mobile data transmission such as the LTE downlink. It is worth noting that the advantage of separating the transmission into multiple narrowband sub-channels cannot itself translate into robustness against time-variant channels if no channel coding is employed [5].

5.4 OFDMA:
Like OFDM, OFDMA (Orthogonal frequency division multiple access) employs multiple closely spaced sub-carriers, but the subcarriers are divided into groups of subcarriers. Each group is named a sub channel. The sub-carriers that form a sub-channel need not be adjacent. In the downlink, a sub channel may be intended for different receivers. In the uplink, a transmitter may be assigned one or more sub-channels. Sub-channelization defines subchannels that can be allocated to subscriber stations depending on the channel conditions and data requirements. Using sub-channelization, within the same time slot a mobile base station can allocate more transmit power to user devices with low SNR and vice-versa. This also save a user device transmit power as it can concentrate power only on certain sub-channels allocated to it [6]. Apart from having certain advantages it could have from OFDM, the OFDMA waveform exhibits very pronounced envelop deviation resulting in a high PAPR (peak to average power ratio). And the signals having high PAPR requires highly linear power amplifiers like class A, class AB etc. to avoid excessive inter modulation distortion. To achieve this linearity, the amplifiers have to operate with a large back off from their peak power, resulting in decreased power efficiency. Another problem with OFDMA is, while up linking there is an introduction of frequency offset among the different terminals that transmit simultaneously, destroying the concept of orthogonality [17].

59

OFDM

Sub-carriers

Time

OFDMA

Sub-channels

Time

Fig 5.4: Difference between OFDM and OFDMA

Consecutive OFDM Subcarriers in Time domain 2

1.8

1.6

Amplitude

1.4

1.2

0.8

5 Subcarrier index

10

15

Fig 5.5: Sensitivity of OFDM subcarriers with Carrier

60

OFDM Transmission Spectrum 1.8 1.6 1.4 1.2

Amplitude

1 0.8 0.6 0.4 0.2 0

50

100

150 Subcarriers

200

250

300

Fig 5.6: OFDM transmission spectrum

5.4.1 Block diagram of OFDMA: As we move ahead for higher generation of mobile technology we always encounter the need of high speed communication. Various multicarrier multiplexing techniques have evolved to meet these demands, some of them being code division multiple access (CDMA) and orthogonal frequency division multiplexing (OFDM). OFDM utilizes orthogonal subcarriers to transmit information parallel. In a conventional serial data transmission, the symbols are transmitted sequentially, with the frequency spectrum of each data symbol allowed to occupy the entire bandwidth. In OFDM, the data is divided among large number of closely spaced carriers (frequency division multiplexing). This is not a multiple access technique, since no common medium is to be shared. Here only small amount of data is carried by each carrier, reducing the ISI significantly. Many modulation schemes could be used to modulate the data at a low bit rate onto each carrier. Bandwidth occupied by the OFDM systems being greater than the correlation bandwidth of the fading channel gives it an extra edge over serial communication [6]. 61

Carrier
Coding & modulation S/P conversion N-IFFT CP insertion P/S conversion

Info Symbol

Channel

Estimated symbol

Decoding & demodulation

P/S conversion

FFT

CP extraction

S/P conversion

Carrier

Fig 5.7: Block Diagram of OFDMA Dividing an entire channel into many narrow sub bands makes the frequency response become relatively flat in each individual sub band. Since each sub channel covers only a small fraction original bandwidth, equalization is quite simple (differential encoding may even make equalization unnecessary) [19]. Use of guard interval, systems reaction to delay spread can be reduced. OFDM can be finally said as a form of multicarrier modulation where its carrier spacing is carefully selected so that each subcarrier is orthogonal to the other subcarriers. The block diagram of OFDMA transmitter & receiver is shown in Fig 5.7.

5.5 Description of Problem Statement:

As its clear from the figure many blocks are common to both OFDMA and SCFDMA. At the input to the transmitter, a baseband modulator transforms the binary input to a multilevel sequence of complex numbers xn in one of several possible modulation formats including quaternary PSK (QPSK), 16-level quadrature amplitude modulation (16-QAM) and 64-QAM etc. Then serial bit stream is converted to parallel bit stream of N data points. The first step is to produce a frequency representation Xk of the input symbols. It then maps each of the N DFT outputs to one of the M (>N) orthogonal subcarriers that can be transmitted, where M=N*Q ,Q is the bandwidth expansion factor of symbol sequence. 62

The mapping can be of two types: 1. LFDMA 2. IFDMA In Localized FDMA each terminal uses a set of adjacent subcarriers to transmit its symbols. Thus the bandwidth of an LFDMA transmission is confined to a fraction of the system bandwidth. In Interleaved FDMA the subcarriers used by a terminal are spread over the entire signal band. Fig 5.8 shows two type of mapping in the frequency domain. There are three terminals, each transmitting symbols on four subcarriers in a system with a total of 12 subcarriers. SCFDMA is better against frequency selective fading because its information is spread across the entire signal band. On the other hand, LFDMA can potentially achieve multi-user diversity in the presence of frequency selective fading if it assigns each user to subcarriers in a portion of the signal band where that user has favorable transmission characteristics [17]. After sub-carrier mapping we get the set of M complex sub-carrier amplitudes X1 frequency domain. Then MDFT is performed to convert them into M time domain signals x m . Each xm then modulates a single frequency carrier and all the modulated symbols are transmitted sequentially.

Terminal 1

Terminal 2

Terminal 3

Interleaved

Localized

Fig 5.8: Sub-carrier mapping for 3 users, 12 sub-carriers and 4 sub-carriers per user.

63

5.6 Mathematical Calculation for PAPR:

Let the data block of length N be represented by a vector X= [X0,X1,.,XN-1]T. Duration of any symbol XK in the set X is T and represents one of the sub-carriers set. As the N subcarriers chosen to transmit the signal are orthogonal, so we can have fn = nf, where nf = 1/NT and NT is the duration of the OFDM data block X. The complex data block for the OFDM signal to be transmitted is given by =
1 =1 2 =0

The PAPR of the transmitted signal is defined as PAPR = 0< |()|2 1 2 0 | |

The cumulative distribution function (CDF) is one of the most regularly used parameters, which is used to measure the efficiency of any PAPR technique. Normally, the complementary CDF (CCDF) is used instead of CDF, which helps us to measure the probability that the PAPR of a certain data block exceeds the given threshold [6]. The CDF of the PAPR of the amplitude of a signal sample is given by; = 1 The CCDF of the PAPR of the data block is desired in our case is to compare outputs of various reduction techniques. This is given by: > = 1 = 1

= 1 (1 )

(5.1)

Fig 5.9 shows PAPR distribution for different numbers of OFDM subcarriers.

64

10

10

P(PAPR>z)------------

N=16 N=32 N=128 N=512 N=2048

10

10

10

10

8 z------------

10

12

14

Fig 5.9: PAPR distribution for different numbers of OFDM subcarriers [3].

5.7 Comparison of PAPR for OFDMA And SCFDMA:

SC-FDMA offers similar performance and complexity as OFDM. However, the main advantage of SC-FDMA is the low PAPR (peak-average-power ratio) of the transmit signal. PAPR is defined as the ratio of the peak power to average power of the transmit signal. As PAPR is a major concern at the user terminals, low PAPR makes the SC-FDMA the preferred technology for the uplink transmission. PAPR relates to the power amplifier efficiency at the transmitter, and the maximum power efficiency is achieved when the amplifier operates at the saturation point. Lower PAPR allows operation of the power amplifier close to saturation resulting in higher efficiency. With higher PAPR signal, the power amplifier operating point has to be backed off to lower the signal distortion, and thereby lowering amplifier efficiency. As SC-FDMA modulated signal can be viewed as a single carrier signal, a pulse shaping filter can be applied to transmit signal to further improve PAPR [18].

5.8 Significance of Pulse Shaping Filter in PAPR Analysis:

In digital communication, pulse shaping is one of the methods of changing the waveform of the transmitted pulse. It helps in limiting the effective bandwidth of the transmission and also the ISI caused by the channel can also be kept in control. Nyquist ISI criterion is the commonly used criterion for evaluation of filters. Examples of pulse-shaping filters are [6]: 65

- Sinc filter - Raised cosine filter - Gaussian filter

5.8.1 Sinc filter: A sinc filter is an idealized filter that removes all frequency components above a given bandwidth, leaves the low frequencies alone and has linear phase. The filter's impulse response is a sinc function in the time domain, and its frequency response is a rectangular function. The impulse response of such a filter is given by [19]; h t = 2Bsinc(2Bt) Where, B = arbitrary cutoff frequency.
Plot of Sinc Filter Transfer Function 1 0.9 0.8 0.7 0.6

h(t)---->

0.5 0.4 0.3 0.2 0.1 0 -5

-4

-3

-2

-1

0 t---->

Fig 5.10: The Transfer Function of Sinc Filter

66

5.8.2 Raised cosine filter: Raised-cosine filter is practical to implement and it is in wide use. It has a parametrisable excess bandwidth, so communication systems can choose a trade-off between a more complex filter and spectral efficiency. The raised-cosine filter is an implementation of a lowpass Nyquist filter, i.e., one that has the property of vestigial symmetry. This means that its spectrum exhibits odd symmetry about 1/2T, where T is the symbol-period of the communications system. Its frequency-domain description is a piecewise function, given by [20]. , 1 = 1 + cos 2 2 0, , 1 2 1 1 + < 2 2

and characterized by two values, the roll-off factor, and T, the reciprocal of the symbol-rate.

Plot the transfer function of Raised Cosine filter 1 alpha=0 alpha=0.5 alpha=1

0.8

0.6

Amplitude

0.4

0.2

-0.2

-0.4

10

15

20 25 n(Samples)

30

35

40

45

Fig 5.11: The Transfer Function of Raised Cosine Filter 67

5.8.3 Gaussian filter: A Gaussian filter is a filter whose impulse response is a Gaussian function. Gaussian filters are designed to give no overshoot to a step function input while minimizing the rise and fall time. This behavior is closely connected to the fact that the Gaussian filter has the minimum possible group delay. The one-dimensional Gaussian filter has an impulse response given by [21]; g x = a a.x 2 .e

Plot of Gaussian Filter Transfer Function 0.2 0.18 0.16 0.14 0.12

g(x)---->

0.1 0.08 0.06 0.04 0.02 0

10 x---->

15

20

25

Fig 5.12: The Transfer Function of Gaussian Filter

5.9 PAPR Reduction Techniques for OFDM Signal:

Many techniques have been studied for reducing the PAPR of a transmitted OFDM signal. Although no such techniques are specied for the LTE downlink signal generation, an overview of the possibilities is provided below. In general in LTE the cost and complexity of generating the OFDM signal with acceptable Error Vector Magnitude (EVM) is left to the 68

eNodeB implementation. As OFDM is not used for the LTE uplink, such considerations do not directly apply to the transmitter in the UE. Techniques for PAPR reduction of OFDM signals [5] can be broadly categorized into three main concepts: Clipping and Filtering Selected Mapping Pre-coding Technique.

5.9.1 Clipping and filtering: The time-domain signal is clipped to a predened level. This causes spectral leakage into adjacent channels, resulting in reduced spectral efficiency as well as in-band noise degrading the bit error rate performance. Out-of- band radiation caused by the clipping process can, however, be reduced by ltering. If discrete signals are clipped directly, the resulting clipping noise will all fall in band and thus cannot be reduced by ltering. To avoid this problem, one solution consists of oversampling the original signal by padding the input signal with zeros and processing it using a longer IFFT. The oversampled signal is clipped and then ltered to reduce the out-of-band radiation. We have chosen a concatenation of interleaving with repeated clipping and filtering using optimum value of and frequency domain filtering. A schematic diagram of the proposed OFDM transmitter is shown in Fig 5.13.

Encoder Input

Interleaving (W)

IFFT (with Over sampling)

Clipping

IFFT

Out-of-Band Removal

FFT

Fig 5.13: Simplified clipping and filtering with Optimum value of

First, the interleaving approach is used and the signal with lowest PAPR is then passed through clipping and filtering method. The intention to combine these two methods is to 69

obtain signal with lower PAPR than in the case of interleaving method and with lower distortion (and thus lower bit error rate) than in the case of standalone Repeated clipping and filtering. As both methods used in the combination suffer from high complexity, the main disadvantage of the combined method is above all the complexity. Moreover, side information (SI) to identify the interleaver with lowest PAPR has to be sent to receiver for each OFDM symbol. Without this side information, it is not possible to decode the data. As the correct decoding of side information is fundamental for the performance of OFDM modem, the side information can thus be either mapped using modulation with lower number of states or encoded by FEC. The complexity of the presented combined method can be dramatically reduced using the recently proposed method Simplified clipping and filtering instead of the repeated clipping and frequency domain filtering method [22]. The clipping and frequency domain filtering of the input OFDM signal is shown in Fig 5.14.

Sc (1) Sc (2) Input Clip OFDM Sc (9) FFT Filtering 0 0 IFFT OFDM Output

Fig 5.14: The clipping and frequency domain filtering of the input OFDM signal.

The modified CF algorithm can be stated as below [22]: 1. Convert the OFDM symbol to time domain as (n) = IFFT (XK). 2. Calculate the optimum value of clipping level and Clip (n) to the threshold A. 3. Convert (n) to frequency domain to obtain Xk by doing FFT of x(n). 4. Clipped the OFDM signal using optimum value and pass through a frequency domain filter based upon Hanning Windowing to reduce the PAPR of OFDM signal. 5. Convert to time domain and transmit the OFDM Signal.

70

5.9.2 Selected mapping: Multiple transmit signals which represent the same OFDM data symbol are generated by multiplying the OFDM symbol by different phase vectors. The representation with the lowest PAPR is selected. To recover the phase information, it is of course necessary to use separate control signaling to indicate to the receiver which phase vector was used [5].
The selective mapping (SLM) technique can be applied to SFBC-OFDM systems with two

transmitter antennas and Almouti coding scheme without changing the orthogonality of space frequency coding. In this method, the optimum phase sequence is applied to the OFDM frames of two antennas such that the SFBC structure remains constant. In the SLM method, D different representations of the OFDM frame are generated, and that with minimum PAPR is transmitted. The main disadvantage of this system is that it increases the complexity of the system by adding a no of terms [23].

Fig 5.15: Block diagram of SFBC-OFDM transmitter with two transmitter antennas and the selective mapping (SLM) method for PAPR reduction

71

5.9.3 Pre-coding technique: These techniques consist of nding the code words with the lowest PAPR from a set of codeword to map the input data. A look-up table may be used if N is small. It is shown that complementary codes have good properties to combine both PAPR and forward error correction. The latter two concepts are not applicable in the context of LTE; selected mapping would require additional signaling, while techniques based on codeword selection are not compatible with the data scrambling used in the LTE downlink [5].

A design procedure for good pre-coding schemes is very important. It is possible to reduce the PAPR of OFDM signals by pre-coding without destroying the detectability property of the different symbols of the OFDM block . We can use any band efficient modulation like BPSK, QPSK etc [24].

Modulation

S/P conversion

Pre coding

IFFT

CP insertion

CFBD

Noise

Channel

Demodulation

P/S conversion

Reverse Pre coding

Turbo Equalizer

FFT

Fig 5.16: Block diagram of pre-coding technique for PAPR reduction of OFDM signal

72

Chapter 6 Characteristics of Mobile Radio Channel

6.1 Introduction:
A channel ideally should contain only one copy of transmitted signal coming in the line of Sight path from transmitter to receiver, so there would be a perfect reconstruction of original signal. But in reality this doesnt happen. Rather the received signal consists of a

combination of attenuated, reflected, refracted and diffracted replicas of original signal . So the channel gets faded both in time and frequency domain. Also the channel adds noise to the signal which further complicates the procedure. If theres relative motion in the channel then frequency shift occurs (Doppler Effect). Knowledge of all these phenomena is necessary in order to model the channel for radio wave propagation [6].

6.2 Types of Fading:

The propagation model mainly focuses on predicting the average received signal strength at a given T-R (Transmitter-Receiver) separation and radial variation for the specified separation. So we can classify fading into two types: Large-scale fading and Small scale fading. Large scale fading attributes for variation in signal strength over large T-R separation distances. Large scale models try to find out mean signal power attenuation or path loss due motion over large area around transmitter or receiver. Small scale fading characterizes rapid

fluctuation of received signal strength over short T-R separations and for short period of time. So the signal is a sum of many signals coming from different directions with different attenuation which brings dramatic changes in signal amplitude and phase. Various models exist in literature for large scale fading. They are like empirical models such as Okumura model, Hata model, cost 136 model etc; indoor models like Log-distance path loss model, Ericsson multiple breakpoint model, Attenuation factor model etc. Large scale fading models find applications in wireless network planning for an area and modeling path loss over a large distance. So, large scale path loss models are more important for cell site planning but less for communication system design. So we will next discuss small scale fading in a little detail.

73

6.3 Small-Scale Fading:

Fading is caused by interference between two or more forms of transmitted signal that arrive at receiver at slightly different times. These components are called multipath components. The complete set of multi paths has to be known for modeling the multipath channel. Each path is characterized by three parameters namely delay, attenuation and phase shift. The discrete time variant channel impulse response of the multipath channel is given by [10]

, =

( ())

where, is the attenuation in the mth path at time t

() is the propagation delay in the mth path at time t 2 is the phase shift for carrier frequency fc for mth multipath component

( ) is the dirac delta function The above model takes into all the modifications that a multipath channel can make to the signal.

6.4 Critical Channel Parameters:

Two kinds of spreading occur when a signal passes through a channel. They are - Multipath delay spread - Doppler spread Multipath delay spread occurs because of time dispersive nature of the channel in local area. Because delayed versions of original signal is superimposed at the receiver so the received signal spreads in time domain or shows time dispersion. Parameter used to describe this is rms delay spread denoted by t. This is defined as the standard deviation value of the delay weighed proportional to the energy of waves. Coherence bandwidth (f0) is analogous to delay spread used to characterize the channel in frequency domain. Its the statistical measure of the range of frequencies for which all components are passed with equal gain and linear phase. So we can say 0
1

74

Doppler spread occurs because of relative motion between transmitter and receiver or motion of objects in the channel. So it occurs because of time variance nature of the channel. Because of relative motion Doppler shift of frequency occurs which broadens the signal in frequency domain or shows frequency dispersion. Parameter used to characterize this is Doppler spread denoted by fd. This is defined as the range of frequencies over which the Doppler spectrum is non-zero. Coherence time (Tc) is the time domain dual of Doppler spread and used to characterize the time varying nature of the frequency depressiveness of channel. Its the statistical measure of time duration over which the channel impulse
1

response is essentially constant. So that we can write

6.5 Types of Small-Scale Fading:

Small-scale fading occurs due to two propagation mechanisms as described above [25]. They are Due to multipath delay spread - Flat fading - Frequency selective fading Due to Doppler spread - Fast fading - Slow fading

If the bandwidth of the channel is less than range of frequency over which the channel has constant gain and linear phase, then the signal undergoes flat fading. This type of fading is common in literature as this is analogous to a low pass filter. After passing the spectral characteristics of the channel remains unchanged but the gain changes with time. So in terms of channel parameters If fm < f0 and Ts > where fm = signal bandwidth and Ts = symbol period

Then the channel creates flat fading If the channel has constant gain and linear phase response over range of frequencies which is less than the signal bandwidth then the channel creates frequency selective fading. Thats different frequency components are faded differently. In time domain the received signal is a distorted because of multiple delayed and faded instances of transmitted signal. As signal gets 75

dispersed in time domain, so channel induces ISI (Inter Symbol Interference). In terms of channel parameters If fm < f0 and Ts > Then the channel creates frequency selective fading In a fast fading channel, the channel characteristics change multiple times within the symbol duration thats it changes at a rate higher than that of the transmitted signal. So this causes frequency dispersion which happens because of high Doppler spreading. We can say low data rate signals have more chance of being fast faded. Thus The signal suffers fast fading if Ts > Tc and fm < fd In a slow fading channel, the channel impulse response change at a rate much lower than that of the transmitted signal. In time domain the channel characteristics remain almost constant during one symbol time. The Doppler spread here is less as compared bandwidth of the baseband signal. Thus The signal suffers fast fading if Ts < Tc and fm < fd So we can say the relative motion between mobile and receiver determines the channel to be slow fading or fast fading.

6.6 Rayleigh and Ricean Distribution:

In a multipath channel if the propagation delays due to multi paths becomes random and the no of multi paths becomes very large, then central limit theorem applies. So the received signal envelope becomes Gaussian and can be modeled using various distribution functions [6].

- Rayleigh distribution: When phase and quadrature component of received envelope are independent and Gaussian with zero mean then the pdf of amplitude assumes Rayleigh distribution. Theres no line of sight path between transmitter and receiver. The power is exponentially distributed. Mostly used as it represents a general case.

76

Fig 6.1 shows an animated effect showing the results of passing an unmodulated carrier through a simple two path Rayleigh fading channel. The animation shows, input (blue) and the output (red), the phase shifts, gains, and attenuations of the output sine wave or carrier.

Rayleigh fading channel with two path sine wave input 2 1.5 1 0.5

amplitude

0 -0.5 -1 -1.5 -2

5 time

10

Fig 6.1: Rayleigh fading channel with two path sine wave input.

- Ricean distribution: Due to deterministic dominant term at least one of in-phase and quadrature component of the received envelope has non-zero mean. So now the pdf of received envelope assumes Ricean distribution. Theres a dominant line of sight path between transmitter and receiver. It applies to microcellular systems.

77

Chapter 7 Channel Estimation in OFDM

7.1 Introduction:
A wideband radio channel is frequency selective and also time variant. In both the frequency and time domain, the channel impulse response at different subcarriers, appear unequal. So we need to estimate the state of the channel at every instant. Pilot based approaches are widely implemented to estimate the channel characteristics and to correct the corrupt channel due to multipath fading. We have basically two kinds of pilot arrangement depending on the nature of channel.

7.2 Block Type of Pilot Arrangement:

The first one, block-type pilot channel estimation, is developed under the assumption of slow fading channel, and is performed by inserting pilot tones into all subcarriers of OFDM symbols within a specific period [6]. As the training block contains all the pilots, channel interpolation in frequency domain is not required. So this type of pilot can be said to be insensitive to frequency selectivity. As the coherence time is higher than the symbol period in slow fading due to lower Doppler spread, the channel characteristics remains almost static for one symbol block time duration. Fig 7.1 shows block-type pilot channel estimation.

7.3 Comb Type of Pilot Arrangement:

The second kind of pilot arrangement is denoted as comb-type pilot arrangement. Assuming the payloads of the pilot arrangements are the same, the comb type pilot arrangement has higher re-transmission rate. Thus the comb-type pilot arrangement gives better resistance to fast fading channels. Since only few subcarriers contain the pilot signal, the channel impulse response of non-pilot sub-carriers can be estimated by either linear, cubic or spline interpolation of the neighboring pilot sub-carriers. We can conclude that such pilot arrangement is sensitive to frequency selectivity. As the coherence time is less than the symbol period in fast fading due to higher spread, the channel characteristics fluctuates many time within one symbol block time duration. Fig 7.1 shows comb-type pilot channel estimation. 78

Block

Pilot Data Frequency

Time Block type pilot channel estimation

Pilot
Frequency Data

Time Comb-type pilot channel estimation

Fig 7.1: Two basic types of pilot arrangement for OFDM channel estimation.

7.4 Working Environment:

As we are taking a Rayleigh slow fading channel we have stressed on the various block type pilot arrangement of channel estimation. In block-type pilot based channel estimation, OFDM channel estimation symbols are transmitted periodically, and all the subcarriers are used as pilots. So in our work we are using a general model for a slowly fading channel, where we

79

make use of MMSE (minimum mean square error) and LS (least square) estimator and a method for modifications compromising between complexity and performance.
X0

Y0

h0

IDF T
YN-1 XN-1

Transfo rmation

DF T
hN-1

Fig 7.2: General estimator structure

The use of DPSK (differential phase shift keying) in OFDM systems avoids the tracking of a time varying channel. However, this limits the number of bits per symbol and results in 3 decibels loss in SNR. If we have a channel estimator in the receiver side, multi amplitude signaling schemes or M-ary PSK modulation schemes can be used. We have worked on BPSK, QPSK, 16 QAM modulation schemes for this purpose. Now we will look into the various estimation techniques in detail and compare the biasedness, complexity and performance of each. Performance is presented both in terms of Mean Square Error (MSE) and Symbol Error Rate (SER). The general estimator structure is shown in Fig 7.2 [26].

7.5 Mathematical Analysis of the Channel Estimators:

Let g : the time domain channel vector h : the frequency domain channel vector X : the diagonal matrix containing mapped input symbols W : white gaussian noise vector Then output symbols in time domain are given by Y = XFg + W = Xh + W Where: X = diag { x0,x1.xN-1} 80

Y= [y0,y1..yN-1]T W=[W0,W1..WN-1]T h = [h0,h1,.hN-1]T = DFTN {g} F = DFT transform block

7.5.1 Least Square Error (LSE) Estimation: The LS estimator minimizes (Y-XFg)H (Y-XFg) w.r.t g Time domain LS estimate of g is given by LS = FHX-1Y LS = X-1Y So Q block in the fig. 4.2 for LS estimator is given by QLS = (FHXHXF)-1 The MMSE estimator suffers from higher complexity because it requires the calculation of an N x N matrix QMMSE, whose complexity increases with increase in N. LS doesnt use any channel statistics, has low complexity but estimator gives higher mean square error.

10

SNR vs BER using LSE estimator for an OFDM Channel

10

-1

Bit Error Rate

10

-2

10

-3

10

-4

10

15 SNR in dB

20

25

30

Fig 7.3: SNR vs BER using LSE estimator for an OFDM channel. 81

So we need to move on to another kind of estimator which would overcome the drawback of both the methods. SNR vs BER using LSE estimator for an OFDM channel is shown in Fig 7.3.

7.5.2 Minimum Mean Square Error (MMSE) Estimation: If the g is uncorrelated with W then the time domain MMSE estimator is given by [26] MMSE = RYg RYY-1Y where RYg : cross-covariance matrix of Y and g RYY : auto-covariance matrix of Y MMSE = FMMSE So Q block in the fig. 4.2 for MMSE estimator is given by QMMSE = Rgg[(FHXHXF)-1n2 + Rgg]-1(FHXHXF)-1 Rgg = auto-covariance matrix of g

SNR vs MSE for an OFDM system with MMSE / LSE estimator has been shown in Fig 7.4. SNR vs SER for an OFDM system with MMSE / LSE estimator has been shown in Fig 7.5.

PLOT OF SNR V/S MSE FOR AN OFDM SYSTEM WITH MMSE/LSE ESTIMATOR BASED RECEIVERS -1 10 MMSE LSE

mean squared error

10

-2

10

-3

10

-4

10

15 SNR in DB

20

25

Fig 7.4: SNR vs MSE for an OFDM system with MMSE / LSE estimator. 82

PLOT OF SNR V/S SER FOR AN OFDM SYSTEM WITH MMSE/LSE ESTIMATOR BASED RECEIVERS MMSE LSE

10

-1.2

Symbol Error Rate

10

-1.3

10

-1.4

10

-1.5

10

15 SNR in DB

20

25

30

Fig 7.5: SNR vs SER for an OFDM system with MMSE / LSE estimator.

7.6 Modified MMSE Estimation:

A straightforward way of decreasing the complexity is to reduce the size of QMMSE. As most of the energy in g is contained in, or near the first L taps as shown in [27] a modification of MMSE estimator, where only the taps with significant energy are considered. The components in Rgg corresponding to low energy taps in g are approximated to zero. So Rgg is an L L matrix containing the covariance of first L components of g. The DFT matrix also needs modification for finding DFT of such matrix. Now it would be an N L matrix by taking only the first L columns of DFT matrix. If T denotes the modified DFT matrix, then as shown in [23] MMSE = TQMMSE THXHY Where, Q MMSE = R gg[(THXHXT)-1n2 + Rgg]-1(THXHXT)-1

As L is a very small fraction of N then the computational burden sharply decreases. As we know the LS estimator doesnt use the statistics of channel only depends on input and output. So modification to LS estimator isnt required as it wont relieve any computational burden.

83

Chapter 8 Simulations & Results

8.1 OFDM Signal and Its Spectrum with Guard Interval:
OFDM signal and its spectrum with Guard Interval -Graph on time domain-

10

10

-1

abs(z1)----->
10
-2

10

-3

-80

-60

-40

-20

0 f----->

20

40

60

80

Fig 8.1: OFDM signal and its spectrum with guard interval (Graph on time domain)

OFDM signal and its spectrum with Guard Interval -Graph on frequency domain20 15 10 5

Y4----->

0 -5 -10 -15 -20

-60

-40

-20

0 f----->

20

40

60

Fig 8.2: OFDM signal and its spectrum with guard interval (Graph on frequency domain)

84

8.2 Comparison of PAPR for OFDMA and SCFDMA for Various Parameters:

Parameters Modulation format Number of total subcarriers (M) Data block size (N) System bandwidth Oversampling factor Number of runs

Values Q-PSK

512

64 5 MHz 4 1000

10

PAPR Analysis for OFDMA & SCFDMA'( N=64 & M= 512) SCFDMA OFDMA

Pr[PAPR>PAPR0]

10

-1

10

-2

10

-3

7 8 PAPR[dB]

10

11

12

Fig 8.3: CCDF of PAPR for OFDMA & SCFDMA (N=64, M=512) with QPSK Modulation 85

Parameters Modulation format Number of total subcarriers (M) Data block size (N) System bandwidth Oversampling factor Number of runs

Values Q-PSK

256

64 5 MHz 4 1000

10

PAPR Analysis for OFDMA & SCFDMA'( N=64 & M= 256) SCFDMA OFDMA

Pr[PAPR>PAPR0]

10

-1

10

-2

10

-3

7 8 PAPR[dB]

10

11

12

Fig 8.4: CCDF of PAPR for OFDMA & SCFDMA (N=64, M=256) with QPSK Modulation

86

Parameters Modulation format Number of total subcarriers. Data block size System bandwidth Oversampling factor Number of runs

Values Q-PSK

128

64 5 MHz 4 1000

10

PAPR Analysis for OFDMA & SCFDMA'( N=64 & M= 128) SCFDMA OFDMA

Pr[PAPR>PAPR0]

10

-1

10

-2

10

-3

7 8 PAPR[dB]

10

11

12

Fig 8.5: CCDF of PAPR for OFDMA & SCFDMA (N=64, M=128) with QPSK Modulation.

87

Parameters Modulation format Number of total subcarriers. Data block size System bandwidth Oversampling factor Number of runs

Values 16-QAM

512

64 5 MHz 4 1000

10

PAPR Analysis for OFDMA & SCFDMA'( N=64& M= 512) SCFDMA OFDMA

Pr[PAPR>PAPR0]

10

-1

10

-2

10

-3

7 PAPR[dB]

10

11

Fig 8.6: CCDF of PAPR for OFDMA & SCFDMA (N=64, M=512) with 16 QAM Modulation.

88

Parameters Modulation format Number of total subcarriers. Data block size System bandwidth Oversampling factor Number of runs

Values 16-QAM

256

64 5 MHz 4 1000

10

PAPR Analysis for OFDMA & SCFDMA'( N=64& M= 256) SCFDMA OFDMA

Pr[PAPR>PAPR0]

10

-1

10

-2

10

-3

7 PAPR[dB]

10

11

Fig 8.7: CCDF of PAPR for OFDMA & SCFDMA (N=64, M=256) with 16 QAM Modulation.

89

Parameters Modulation format Number of total subcarriers. Data block size System bandwidth Oversampling factor Number of runs

Values 16-QAM

128

64 5 MHz 4 1000

10

PAPR Analysis for OFDMA & SCFDMA'( N=64& M= 128) SCFDMA OFDMA

Pr[PAPR>PAPR0]

10

-1

10

-2

10

-3

7 PAPR[dB]

10

11

Fig 8.8: CCDF of PAPR for OFDMA & SCFDMA (N=64, M=128) with 16 QAM Modulation.

90

Parameters Modulation format Number of total subcarriers. Data block size System bandwidth Oversampling factor Number of runs

Values 16-QAM

128

16 5 MHz 4 1000

10

PAPR Analysis for OFDMA & SCFDMA'( N=16& M= 128) SCFDMA OFDMA

Pr[PAPR>PAPR0]

10

-1

10

-2

10

-3

6 PAPR[dB]

10

Fig 8.9: CCDF of PAPR for OFDMA & SCFDMA (N=16, M=128) with 16 QAM Modulation.

91

The above figures clearly shows that the CCDF of PAPR for OFDMA signal contains high PAPR and the CCDF of PAPR for SCFDMA signal contains low PAPR. PAPR comparison between OFDMA and SC-FDMA showed that low PAPR makes the SC-FDMA the most preferred technology for the uplink transmission in LTE system.

8.3 Investigation of Clipping & Filtering Method as PAPR Reduction Technique for OFDM Signals:
PERFORMANCE CHARACTERISTICS: Transmitted Data Phase Representation: The data phase representation of 1*128 randomly generated data points in the transmitter section is shown above.
Transmitted Data "O" 3

transmitted data phase representation

2.5

1.5

0.5

20

40

60 80 Data Points

100

120

140

Fig 8.10: Transmitted Data Phase Representation

Scatter Plot : Scatter plot help in the representation of the modulated signal in the signal space by plotting its in-phase components against its quadrature phase.

92

M represents the alphabet size and must be an integer power of 2.Since QPSK modulation technique is being applied therefore the above figure shows the M = 4, i.e, quadrature phase components.
MODULATED TRANSMITTED DATA 1 0.8 0.6 0.4

Quadrature

0.2 0 -0.2 -0.4 -0.6 -0.8 -1 -1 -0.5 0 In-Phase 0.5 1

Fig 8.11: The representation of the modulated signal (QPSK)

93

Time versus Amplitude plot: In clipped OFDM signal shown above, the amplitude varies from -0.4 to +0.4 whereas in unclipped OFDM signal it may exceed this average value.
Unclipped OFDM Signal 0.6

0.4

0.2

Amplitude

-0.2

-0.4

-0.6

-0.8

50 Time

100

150

Fig 8.12: Unclipped OFDM signal

clipped OFDM Signal 0.4 0.3 0.2 0.1

Amplitude

0 -0.1 -0.2 -0.3 -0.4

50 Time

100

150

Fig 8.13: Clipped OFDM signal

94

Unclipped and clipped OFDM signal after passing through high power amplifier: To show the effect of High Power Amplifier (H.P.A.), random complex noise is added when the power exceeds the average value ,i.e., -0.4 to +0.4 otherwise no addition is done.
Unclipped OFDM Signal after HPA 2

Amplitude

-1

-2

-3

-4

50 Time

100

150

Fig 8.14: Unclipped OFDM signal after passing through H.P.A

clipped OFDM Signal after HPA 0.4 0.3 0.2 0.1

Amplitude

0 -0.1 -0.2 -0.3 -0.4

50 Time

100

150

Fig 8.15: Clipped OFDM signal after passing through H.P.A

95

Comparison between data phase representation of transmitted OFDM signal received unclipped OFDM signal: When the data phase representation of the transmitted OFDM signal is compared to that of data phase representation of the received unclipped OFDM signal, it is found that only few data points gets matched.
Transmitted Data "O" 3

transmitted data phase representation

2.5

1.5

0.5

20

40

60 80 Data Points

100

120

140

Received Unclipped OFDM Signal "X" 3

2.5

received data phase representation

1.5

0.5

20

40

60 80 Data Points

100

120

140

Fig 8.16: Comparison between Transmitted Data Phase Representation & Received unclipped OFDM signal

96

Comparison between data phase representation of transmitted OFDM signal received clipped OFDM signal: When the data phase representation of the transmitted OFDM signal is compared to that of data phase representation of the received clipped OFDM signal ,it is found that large number of data points gets matched.
Transmitted Data "O" 3

transmitted data phase representation

2.5

1.5

0.5

20

40

60 80 Data Points

100

120

140

Received clipped OFDM Signal "X" 3

2.5

received data phase representation

1.5

0.5

20

40

60 80 Data Points

100

120

140

Fig 8.17: Comparison between Transmitted Data Phase Representation & Received clipped OFDM signal 97

This investigation proposes a novel CLIPPING scheme for the reduction of PAPR, when data phase representation of unclipped OFDM signal and clipped OFDM signal is compared with the data phase representation of transmitted OFDM signal then it can be concluded that PAPR get reduced in clipped OFDM signal. Thus, an improved signal is obtained.

98

Conclusion & Future Scope

SC-FDMA offers similar performance and complexity as OFDM. In this thesis, PAPR comparison between OFDMA and SC-FDMA showed that PAPR is a major concern at the user terminals, low PAPR makes the SC-FDMA the most preferred technology for the uplink transmission. PAPR relates to the power amplifier efficiency at the transmitter, and the maximum power efficiency is achieved when the amplifier operates at the saturation point. Lower PAPR allows operation of the power amplifier close to saturation resulting in higher efficiency. With higher PAPR signal, the power amplifier operating point has to be backed off to lower the signal distortion, and thereby lowering amplifier efficiency. This thesis also investigated the effects of high power amplifier and the channel noise on the OFDM signals and then introduces clipping and filtering as a PAPR reduction method to reduce the PAPR. As SC-FDMA modulated signal can be viewed as a single carrier signal, a pulse shaping filter can be applied to transmit signal to further improve PAPR in Future. Other PAPR reduction techniques for OFDMA can be used in next and compared the techniques with each other by means of various factors (such as performance, system configuration or implementation costs and complexity) to select the better one for practical implementation.

99

References
[1] http://en.wikipedia.org/wiki/LTE_(telecommunication) [2] http://compnetworking.about.com/od/cellularinternetaccess/g/lte-broadband.htm [3] http://www.mobileburn.com/definition.jsp?term=LTE [4] http://www.3gpp.com/About-3GPP [5] S. Sesia, I. Touk, and M. Baker, LTE, The UMTS Long Term Evolution: From Theory to Practice. John Wiley & Sons, 2009. [6] A thesis paper on PAPR ANALYSIS AND CHANNEL ESTIMATION TECHNIQUES FOR 3GPP LTE SYSTEM by By Abhijeet Sahu & Soumyajyoti Behera. [7] http://en.wikipedia.org/wiki/1G [8] http://en.wikipedia.org/wiki/2G [9] http://en.wikipedia.org/wiki/3G [10] http://en.wikipedia.org/wiki/4G, [11] http://wireless.agilent.com/wireless/helpfiles/n7624a/sc-fdma_frame_structure.htm [12] http://0el70lte.wordpress.com/2012/06/25/hello-world/ [13] http://en.wikipedia.org/wiki/Single-carrier_FDMA [14] A project on CHANNEL ESTIMATION AND PREDICTION IN UMTS LTE, Aalborg University, Institute of Electronic Systems Signal and Information Processing for Communications, 2007. [15] Performance Evaluation of LTE Physical Layer Using SC-FDMA & OFDMA By Abdul Samad Shaikh & Khatri Chandan Kumar, Blekinge Institute of Technology, School of Engineering, Department of Radio Communication, Nov 2010. [16] PAPR REDUCTION IN OFDM COMMUNICATIONS WITH GENERALIZED DISCRETE FOURIER TRANSFORM, By Sertac Sayin. [17] Hyung G. Myung, Junsung Lim, and David J. Goodman, Single Carrier FDMA for Uplink Wireless Transmission, IEEE Vehicular Technology Magazine, vol. 1, no. 3, Sep. 2006, pp. 3038. [18] Single Carrier FDMA in LTE, IXIA, Rev A Nov, 2009. [19] http://en.wikipedia.org/wiki/Sinc_filter [20] http://en.wikipedia.org/wiki/Raised-cosine_filter [21] http://en.wikipedia.org/wiki/Gaussian_filter [22] MODIFIED CLIPPING AND FILTERING TECHNIQUE FOR PEAK-TO-AVERAGE POWER RATIO REDUCTION OF OFDM SIGNALS USED IN WLAN by P.K.Sharma, Department of Electronics and Communication Engineering, Bhagwan Parshuram Institute of Technology, GGSIP University, Delhi, India. Vol. 2(10), 2010.

100

[23] SELECTED MAPPING ALGORITHM FOR PAPR REDUCTION OF SFBC OFDM BY ADDING DCT by Nisharaj.R.S, P. Thiruvalar Selvan. [24] Improved Precoding Method for PAPR Reduction in OFDM with Bounded Distortion by Namitha.A.S & Sudheesh.P, International Journal of Computer Applications (0975 8887), June 7, 2010.
[25] Suhas Mathur, Small Scale Fading in Radio Propagation, Department of Electrical Engineering, Rugters University, Lecture Notes for Wireless Communication Technologies, Spring 2005 [26] O.Edfors, M.Sandell, J.Beek, S. K.Wilson, and P. O. Borjesson, OFDM channel estimation by singular value decomposition, IEEE Transaction on Communications, vol. 46, no. 7, pp.931-939, July 1998. [27] J. V. de Beek, O. Edfors, M. Sandel, S. Wilson, and P. Borjesson, On channel estimation in ofdm systems, in IEEE 45th Vehicular Technology Conference, Chicago, USA , Jul. 1995.

101

Appendix A MATLAB Codes Used for PAPR Analysis

1. SCFDMA PAPR Simulation Matlab Code For QPSK Modulation:

% SCFDMA PAPR Simulation Matlab Code For QPSK Modulation% % Mahmud --------------% function papr(input) totalSubcarriers = 512; % Number of total subcarriers. numSymbols = 64; % Data block size. numRuns = 1000 Fs = 5e6; % System bandwidth. Ts = 1/Fs; % System sampling rate. Nos = 4; % Oversampling factor.; papr=zeros(numRuns,1); table=ones(400,64); input=zeros(numSymbols,1); color=['r'] k=1; for n = 1:numRuns, % Generate random data tmp = round(rand(numSymbols,2)); tmp = tmp*2 - 1; data = (tmp(:,1) + 1i*tmp(:,2))/sqrt(2); X = fft(data); X=X.*exp(1i*pi*input); Y = zeros(totalSubcarriers,1); Y(1:numSymbols) = X; y = ifft(Y); papr(n) =10*log10(max(abs(y).^2) / mean(abs(y).^2)); %-------TO CREATE TABLE----------table(k,:)=data; k=k+1; %--------------------------------end save table table [N,X] = hist(papr, 100); semilogy(X,1-cumsum(N)/max(cumsum(N)),color) title(['PAPR Analysis for OFDMA & SCFDMA''( N=' num2str(numSymbols) ' & M= ' num2str(totalSubcarriers),')']); ylabel('Pr[PAPR>PAPR0]'); xlabel('PAPR[dB]') legend('SCFDMA','OFDMA') grid on hold all

102

2. OFDMA PAPR Simulation Matlab Code For QPSK Modulation:

% OFDMA PAPR Simulation Matlab Code For QPSK Modulation% % Mahmud --------------% function paprOFDMA() dataType = 'Q-PSK'; % Modulation format. totalSubcarriers = 512; % Number of total subcarriers. numSymbols = 64; % Data block size. Fs = 5e6; % System bandwidth. Ts = 1/Fs; % System sampling rate. Nos = 4; % Oversampling factor. Nsub = totalSubcarriers; Fsub = [0:Nsub-1]*Fs/Nsub; % Subcarrier spacing. numRuns = 1000; % Number of runs. papr = zeros(1,numRuns); % Initialize the PAPR results. for n = 1:numRuns, % Generate random data. if dataType == 'Q-PSK' tmp = round(rand(numSymbols,2)); tmp = tmp*2 - 1; data = (tmp(:,1) + j*tmp(:,2))/sqrt(2); for k = 1:numSymbols, if tmp(k) == 0 tmp(k) = 1; end data(k) = dataSet(tmp(k)); end data = data.'; end % Time range of the OFDM symbol. t = [0:Ts/Nos:Nsub*Ts]; % OFDM modulation. y = 0; for k = 1:numSymbols, y= y + data(k)*exp(j*2*pi*Fsub(k)*t); end % Calculate PAPR. papr(n) = 10*log10(max(abs(y).^2) / mean(abs(y).^2)); end % Plot CCDF. [N,X] = hist(papr, 100); semilogy(X,1-cumsum(N)/max(cumsum(N)),'b') title(['PAPR Analysis for OFDMA & SCFDMA''( N=' num2str(numSymbols) ' & M= ' num2str(totalSubcarriers),')']); ylabel('Pr[PAPR>PAPR0]'); xlabel('PAPR[dB]') grid on % Save data. save paprOFDMA legend('SCFDMA','OFDMA') grid on hold all

103

3. SCFDMA PAPR Simulation Matlab Code For 16-QAM Modulation:

% SCFDMA PAPR Simulation Matlab Code For 16-QAM Modulation % % Mahmud-----------% function papr(input) totalSubcarriers = 512; % Number of total subcarriers. numSymbols = 64; % Data block size. Fs = 5e6; % System bandwidth. Ts = 1/Fs; % System sampling rate. Nos = 4; % Oversampling factor ; numRuns = 1000; papr=zeros(numRuns,1); table=ones(numRuns,numSymbols); %---- to see the original performance-input=zeros(numSymbols,1); color=['b']; %------------------------------------sertac=1; for n = 1:numRuns, % Generate random data.a data=ones(1,numSymbols); dataSet = [-3+3i -1+3i 1+3i 3+3i ... -3+1i -1+1i 1+1i 3+1i ... -3-1i -1-1i 1-1i 3-1i ... -3-3i -1-3i 1-3i 3-3i]; dataSet = dataSet / sqrt(mean(abs(dataSet).^2)); tmp = ceil(rand(numSymbols,1)*16); for k = 1:numSymbols, if tmp(k) == 0 tmp(k) = 1; end data(1,k) = dataSet(1,tmp(k)); end data = data.'; X = fft(data); X=X.*exp(1i*pi*input); Y = zeros(totalSubcarriers,1); Y(1:numSymbols) = X; y = ifft(Y); papr(n) =10*log10(max(abs(y).^2) / mean(abs(y).^2)); %-------TO CREATE TABLE table(sertac,:)=data; sertac=sertac+1; end save table table [N,X] = hist(papr, 100); semilogy(X,1-cumsum(N)/max(cumsum(N)),color) title(['PAPR Analysis for OFDMA & SCFDMA''( N=' num2str(numSymbols) '& M= ' num2str(totalSubcarriers),')']); ylabel('Pr[PAPR>PAPR0]'); xlabel('PAPR[dB]') legend('SCFDMA','OFDMA') grid on hold on

104

4. OFDMA PAPR Simulation Matlab Code For 16-QAM Modulation:

% OFDMA PAPR Simulation Matlab Code For 16-QAM Modulation% % Mahmud --------------% function paprOFDMA() dataType = 'Q-PSK'; % Modulation format. totalSubcarriers = 512; % Number of total subcarriers. numSymbols = 64; % Data block size. Fs = 5e6; % System bandwidth. Ts = 1/Fs; % System sampling rate. Nos = 4; % Oversampling factor. Nsub = totalSubcarriers; Fsub = [0:Nsub-1]*Fs/Nsub; % Subcarrier spacing. numRuns = 1000; % Number of runs. papr = zeros(1,numRuns); % Initialize the PAPR results. for n = 1:numRuns, % Generate random data. tmp = round(rand(numSymbols,2)); tmp = tmp*2 - 1; data = (tmp(:,1) + j*tmp(:,2))/sqrt(2); if dataType == '16QAM' dataSet = [-3+3i -1+3i 1+3i 3+3i ... -3+i -1+i 1+i 3+i ... -3-i -1-i 1-i 3-i ... -3-3i -1-3i 1-3i 3-3i]; dataSet = dataSet / sqrt(mean(abs(dataSet).^2)); tmp = ceil(rand(numSymbols,1)*16); for k = 1:numSymbols, if tmp(k) == 0 tmp(k) = 1; end data(k) = dataSet(tmp(k)); end data = data.'; end % Time range of the OFDM symbol. t = [0:Ts/Nos:Nsub*Ts]; % OFDM modulation. y = 0; for k = 1:numSymbols, y= y + data(k)*exp(j*2*pi*Fsub(k)*t); end % Calculate PAPR. papr(n) = 10*log10(max(abs(y).^2) / mean(abs(y).^2)); end % Plot CCDF. [N,X] = hist(papr, 100); semilogy(X,1-cumsum(N)/max(cumsum(N)),'m')

105

title(['PAPR Analysis for OFDMA & SCFDMA''( N=' num2str(numSymbols) '& M= ' num2str(totalSubcarriers),')']); ylabel('Pr[PAPR>PAPR0]'); xlabel('PAPR[dB]') legend('SCFDMA','OFDMA') grid on % Save data. save paprOFDMA hold all;

106

Appendix B Clipping & Filtering Method

OFDM can be seen as either a modulation technique or a multiplexing technique. It uses the phenomenon of multicarrier propagation and hence proves to be an important technique for the transmission of high bit rate data in a radio environment. It provides both TDMA and FDMA and in it a single channel is further subdivided into a number of sub-channels or subcarriers so that multiple data bit streams can be sent in parallel simultaneously without significant losses. Increasing robustness against frequency selective fading or narrowband interference is one of the most important reasons for the popularity of OFDM. However OFDM signal suffers from high PAPR or crest factor which might require a large amplifier power back off. Hence our result oriented investigation show that clipping can improve the PAPR of OFDM signal transmission.

2. System Description: Our investigation with the help of MATLAB CODING (m-file) depends on the analysis of the various subsections as stated below:A: Parameter specifications B: Transmitter section C: Clipping as a PAPR reduction method D: Analyzing of effect of high power amplifier E: Generation of complex multipath channel F: Receiver section The detailed analysis of these sections is being listed below: A: Parameter Specifications: In this section we have assumed an OFDM signal with following specifications: - QPSK signal constellation i.e. M=4; - No _of _data_points=128; - Size of each OFDM block i.e. block_size=8; 107

- Length of cyclic prefix i.e.cp_length=ceil(0.1*block_size); Note:-where ceil rounds the element to the nearest integer towards infinity. - no_of _ifft _points and no_of_fft_points is considered to be equal toblock _size . - Clipping of transmitted signal is done so that a signal remains between +0.4 to -0.4 average value. B: Transmitter Section: Initially 1*128 random data points are generated and then QPSK modulation technique is performed which provides the complex envelope of modulating the message signal using the phase shift keying. Message signal consists of integer values between zero (0) to M-1. Inverse Fast Fourier Transform (IFFT) is now performed on each block by finding out the number of columns that will exist after reshaping an empty matrix is created to put the IFFT data and it operates column wise by appending cyclic prefix which leads to the creation of OFDM block. Data is converted to serial stream for the purpose of transmission and actual OFDM signal to be transmitted is generated.

C: Clipping as a PAPR Reduction Method: OFDM signal suffers from high PAPR or crest factor which may require a large amplifier power back-off. Hence, clipping of transmitted signal is done so that a signal remains between +0.4 to -0.4 average value. D: Analyzing of Effect of High Power Amplifier: In order to show the effect of power amplifier, random complex noise is generated and then clipped signal and original OFDM signal (unclipped) are passed through high power amplifier. E: Generation of Complex Multipath Channel: The signals are transmitted through complex multipath channel to the receiver for the purpose of demodulation. F: Receiver Section: In the receiver section clipped and unclipped data is converted back to parallel form in order to perform Fast Fourier Transform (FFT). Cyclic prefix is removed and data is again converted to serial stream and demodulated.

108

MATLAB codes for PAPR reduction Clipping & Filtering -:

%------Mahmud------------% clear all clc close % --------------% A: Setting Parameters % --------------M = 4; % QPSK signal constellation no_of_data_points = 128; % have 128 data points block_size = 8; % size of each ofdm block cp_len = ceil(0.1*block_size); % length of cyclic prefix no_of_ifft_points = block_size; % 128 points for the FFT/IFFT no_of_fft_points = block_size; % --------------------------------------------% B: % +++++ TRANSMITTER +++++ % --------------------------------------------% 1. Generate 1 x 128 vector of random data points data_source = randsrc(1, no_of_data_points, 0:M-1); figure(1) stem(data_source); grid on; xlabel('Data Points'); ylabel('transmitted data phase representation') title('Transmitted Data "O"')

% 2. Perform QPSK modulation qpsk_modulated_data = pskmod(data_source, M); scatterplot(qpsk_modulated_data);title('MODULATED TRANSMITTED DATA'); % 3. Do IFFT on each block % Make the serial stream a matrix where each column represents a pre-OFDM % block (w/o cyclic prefixing) % First: Find out the number of colums that will exist after reshaping num_cols=length(qpsk_modulated_data)/block_size; data_matrix = reshape(qpsk_modulated_data, block_size, num_cols); % Second: Create empty matix to put the IFFT'd data cp_start = block_size-cp_len; cp_end = block_size; % Third: Operate columnwise & do CP for i=1:num_cols, ifft_data_matrix(:,i) = ifft((data_matrix(:,i)),no_of_ifft_points); % Compute and append Cyclic Prefix for j=1:cp_len, actual_cp(j,i) = ifft_data_matrix(j+cp_start,i); end % Append the CP to the existing block to create the actual OFDM block ifft_data(:,i) = vertcat(actual_cp(:,i),ifft_data_matrix(:,i)); end % 4. Convert to serial stream for transmission [rows_ifft_data cols_ifft_data]=size(ifft_data); len_ofdm_data = rows_ifft_data*cols_ifft_data; % Actual OFDM signal to be transmitted ofdm_signal = reshape(ifft_data, 1, len_ofdm_data); figure(3)

109

plot(real(ofdm_signal)); xlabel('Time'); ylabel('Amplitude'); title('Unclipped OFDM Signal');grid on; % --------------------------------------------------------------% C: % +++++ clipping as a PAPR reduction method +++++ % --------------------------------------------------------------avg=0.4; clipped=ofdm_signal; for i=1:length(clipped) if clipped(i) > avg clipped(i) = avg; end if clipped(i) < -avg clipped(i) = -avg; end end figure(4) plot(real(clipped)); xlabel('Time'); ylabel('Amplitude'); title('clipped OFDM Signal');grid on; % -----------------------------------------% D: % +++++ HPA +++++ % -----------------------------------------%To show the effect of the PA simply we will add random complex noise %when the power exceeds the avg. value, otherwise it add nothing. % 1. Generate random complex noise noise = randn(1,len_ofdm_data) + sqrt(-1)*randn(1,len_ofdm_data); % 2. Transmitted OFDM signal after passing through HPA %without clipping for i=1:length(ofdm_signal) if ofdm_signal(i) > avg ofdm_signal(i) = ofdm_signal(i)+noise(i); end if ofdm_signal(i) < -avg ofdm_signal(i) = ofdm_signal(i)+noise(i); end end figure(5) plot(real(ofdm_signal)); xlabel('Time'); ylabel('Amplitude'); title('Unclipped OFDM Signal after HPA');grid on; %with clipping avg=0.4; for i=1:length(clipped) if clipped(i) > avg clipped(i) = clipped(i)+noise(i); end if clipped(i) < -avg clipped(i) = clipped(i)+noise(i); end end figure(6) plot(real(clipped)); xlabel('Time'); ylabel('Amplitude'); title('clipped OFDM Signal after HPA');grid on; % % -------------------------------E: % +++++ CHANNEL +++++

110

% -------------------------------% Create a complex multipath channel channel = randn(1,block_size) + sqrt(-1)*randn(1,block_size); % % % -----------------------------------------F: % +++++ RECEIVER +++++ ------------------------------------------

% 1. Pass the ofdm signal through the channel after_channel = filter(channel, 1, ofdm_signal); % 2. Add Noise awgn_noise = awgn(zeros(1,length(after_channel)),0); % 3. Add noise to signal...

recvd_signal = awgn_noise+after_channel; % 4. Convert Data back to "parallel" form to perform FFT recvd_signal_matrix = reshape(recvd_signal,rows_ifft_data, cols_ifft_data); % 5. Remove CP recvd_signal_matrix(1:cp_len,:)=[]; % 6. Perform FFT for i=1:cols_ifft_data, % FFT fft_data_matrix(:,i) = fft(recvd_signal_matrix(:,i),no_of_fft_points); end % 7. Convert to serial stream recvd_serial_data = reshape(fft_data_matrix, 1,(block_size*num_cols)); % 8. Demodulate the data qpsk_demodulated_data = pskdemod(recvd_serial_data,M); figure(7) stem(qpsk_demodulated_data,'rx'); grid on;xlabel('Data Points');ylabel('received data phase representation');title('Received Unclipped OFDM Signal "X"') % ---------------------------------------------------% F: % +++++ RECEIVER of clipped signal +++++ % ---------------------------------------------------% 1. Pass the ofdm signal through the channel after_channel = filter(channel, 1, clipped); % 2. Add Noise awgn_noise = awgn(zeros(1,length(after_channel)),0); % 3. Add noise to signal...

recvd_signal = awgn_noise+after_channel; % 4. Convert Data back to "parallel" form to perform FFT recvd_signal_matrix = reshape(recvd_signal,rows_ifft_data, cols_ifft_data); % 5. Remove CP

111

recvd_signal_matrix(1:cp_len,:)=[]; % 6. Perform FFT for i=1:cols_ifft_data, % FFT fft_data_matrix(:,i) = fft(recvd_signal_matrix(:,i),no_of_fft_points); end % 7. Convert to serial stream recvd_serial_data = reshape(fft_data_matrix, 1,(block_size*num_cols)); % 8. Demodulate the data qpsk_demodulated_data = pskdemod(recvd_serial_data,M); figure(8) stem(qpsk_demodulated_data,'rx'); grid on;xlabel('Data Points');ylabel('received data phase representation');title('Received clipped OFDM Signal "X"')

112

Appendix C MATLAB Codes Used for Equations & OFDM spectrum

1. MATLAB code for PAPR eqn for different subcarriers (Eqn 5.1):
% PAPR analysis and simulation for 3GPP LTE system % % Author: S.M.Mahmud Hasan % % This program plots PAPR eqn for different subcarriers % %-----------------------------------------------------% x=2:2:14 N=[16] y=1-(1-exp(-x)).^N semilogy(x,y.^-1) hold all grid on xlabel('z------------') ylabel('P(PAPR>z)------------') x=2:2:14 N=[32] y=1-(1-exp(-x)).^N semilogy(x,y.^-1) hold all grid on xlabel('z------------') ylabel('P(PAPR>z)------------') x=2:2:14 N=[128] y=1-(1-exp(-x)).^N semilogy(x,y.^-1) hold all grid on xlabel('z------------') ylabel('P(PAPR>z)------------') x=2:2:14 N=[512] y=1-(1-exp(-x)).^N semilogy(x,y.^-1) hold all grid on xlabel('z------------') ylabel('P(PAPR>z)------------') x=2:2:14 N=[2048] y=1-(1-exp(-x)).^N semilogy(x,y.^-1) hold all grid on xlabel('z------------') ylabel('P(PAPR>z)------------') legend('N=16','N=32','N=128','N=512','N=2048')

113

2. MATLAB code for the TF of Sinc Filter:

x = linspace(-5,5); y = sinc(x).^2; plot(x,y) title('Plot of Sinc Filter Transfer Function') xlabel('t---->') ylabel('h(t)---->') grid on

3. MATLAB code for the TF of Raised Cosine Filter:

% PAPR analysis and simulation for 3GPP LTE system % % Author: S.M.Mahmud Hasan % % This program plots the Raised Cosine Filter TF % %------------------------------------------------% L=41; %Filter Length R=1E6; %Data Rate = 1Mbps Fs=8*R; %Oversampling by 8 T=1/R; Ts=1/Fs; alpha =0; % Design Factor for Raised Cosing Filter %---------------------------------------------------------%Raised Cosing Filter Design %---------------------------------------------------------if mod(L,2)==0 M=L/2 ; % for even value of L else M=(L-1)/2; % for odd value of L end g=zeros(1,L); %Place holder for RC filter's transfer function for n=-M:M num=sin(pi*n*Ts/T)*cos(alpha*pi*n*Ts/T); den=(pi*n*Ts/T)*(1-(2*alpha*n*Ts/T)^2); g(n+M+1)=num/den; if (1-(2*alpha*n*Ts/T)^2)==0 g(n+M+1)=pi/4*sin(pi*n*Ts/T)/(pi*n*Ts/T); end if n==0 g(n+M+1)=cos(alpha*pi*n*Ts/T)/(1-(2*alpha*n*Ts/T)^2); end end %---------------------------------------------------------% Plot the transfer function of RC filter plot(g); title('Plot the transfer function of Raised Cosine filter') xlabel('n(Samples)'); ylabel('Amplitude'); grid on; hold all; legend('alpha=0','alpha=0.5','alpha=1')

114

4. MATLAB code for the TF of Gaussian Filter:

% PAPR analysis and simulation for 3GPP LTE system % % Author: S.M.Mahmud Hasan % sigma=2; X=-10:10; GAUSS=1/(sqrt(2*pi)*sigma)*exp(-0.5*X.^2/(sigma^2)); plot(GAUSS) title('Plot of Gaussian Filter Transfer Function') xlabel('x---->') ylabel('g(x)---->') grid on

5. MATLAB code for plotting sensitivity of OFDM subcarriers with Carrier:

%% This program plots sensitivity of OFDM subcarriers with Carrier %% frequency offset(CFO) % PAPR analysis and simulation for 3GPP LTE system % % Author: S.M.Mahmud Hasan % clc clear all e = 0; % Normalized CFO N = 16; % Total Subcarriers Indx = 0.01; % Over sampling index vi = 1; % counter index for k = 0:Indx:N-1 hi = 1; % counter index for l = 0:N-1 % this function calculates effect of CFO. Bias 1 is deliberately % added in order to evaluate function at zero CFO. f(vi,hi) = 1 +(sin(pi*(l+e-k))*exp(1i*pi*(N-1)*(l+e-k)/N))... /(N*sin(pi*(l+e-k)/N)); hi = hi+1; end vi = vi+1; end plot([0:Indx:N-1],abs(f(:,1)),'r'); hold on; grid on; title('Consecutive OFDM Subcarriers in Time domain'); xlabel('Subcarrier index');ylabel('Amplitude'); for n = 1:N-1 plot([0:Indx:N-1],abs(f(:,n+1))); end

6. MATLAB code for plotting ofdm trasmission spectrum:

%ofdm trasmission spectrum % PAPR analysis and simulation for 3GPP LTE system % % Author: S.M.Mahmud Hasan (etemahmud@gmail.com)% clc N = 16; % Number of subcarriers. a = sign(randn(N, 1)); % Generate BPSK symbols. b = diag(a); % This helps to plot overlapping subcarrier spectrum. c = ifft(b); % Do IFFT along each column--(each column is a subcarrier). f = fft(c, N*16); % Do FFT of 16x resolution.

115

plot(abs(f)); % U get the spectrum corresponding to each subcarrier. grid on; hold on; title('OFDM Transmission Spectrum'); xlabel('Subcarriers');ylabel('Amplitude'); plot(abs(sum(f, 2)), '-*');

116