Professional Documents
Culture Documents
Unit – 3
G. Roy Antony Arnold
Asst Prof /CSE
Asst. Prof./CSE
GRAA
• IIn contrast
t t tto R l i h which
Rayleigh, hi h models
d l th
the d f t pattern
defect tt off
the entire development process, reliability growth models
are usually based on data from the formal testing phases.
• Indeed it makes more sense to apply these models during
the final testing phase when development is virtually
complete,
complete especially when the testing is customer
oriented.
• During such post‐development testing, when defects are
id tifi d and
identified d fixed,
fi d the
th software
ft b
becomes more stable,
t bl
and reliability grows over time. Therefore models that
address such a process are called
.
GRAA
• They
h are classified
l ifi d into
i two classes.
l They
h are,
– Time between Failure Model
• the variable under study is the time between failures
• Mean time to next failure is usually the parameter to
b estimated
be i d for
f the
h model.
d l
– Fault Count Model
• the
h variable
i bl criterion
i i i the
is h number b off faults
f l or
failures (or normalized rate) in a specified time
interval.
• The number of remaining defects or failures is the key
parameter to be estimated from this class of models.
GRAA
• There are N unknown software faults
at the start of testingg
• Failures occur randomly
• All faults
f l contribute
ib equally
ll to failure
f il
• Fix time is negligibly
g g y small
• Fix is perfect for each fault
GRAA
• Jelinski‐Moranda
J li ki M d (J‐M) Model
(J M) M d l
– Assumes random failures, perfect zero time fixes, all
f l
faults equally bad
ll b d
• Littlewood Models
– Like J‐M model, but assumes bigger faults
found first
• Goel‐Okumoto Imperfect Debugging Model
– Like J
Like J‐M
M model, but with bad fixes possible
model, but with bad fixes possible
GRAA
( )
• One of the earliest model. (1972)
• The software product’s failure rate improves by the same
amount at each fix.
• The hazard function at time ti, the time between the (i‐1)st
and ith failures, is given
GRAA
• Similar to J‐M Model, except it assumes that
y
different faults have different sizes, thereby
contributing unequally to failures. (1981)
• Larger‐sized faults tend to be detected and
Larger sized faults tend to be detected and
fixed earlier.
• This concept makes the model assumption
more realistic.
more realistic.
GRAA
• JJ‐M Model assumes perfect debugging. But this is not
MM d l f t d b i B t thi i t
possible always.
• In the process of fixing a defect, new defects may be
In the process of fixing a defect new defects may be
injected. Indeed, defect fix activities are known to be
error‐prone.
• Hazard function is,
GRAA
• Testing intervals are independent of each
other
• Testing during intervals is reasonably
homogeneous
• Number of defects detected is independent
of each other
GRAA
• G
Goel‐Okumoto
l Ok t Non‐homogeneous Poisson Process
N h P i P
Model (NHPP)
– ## of failures in a time period, exponential failure rate (i.e.
of failures in a time period, exponential failure rate (i.e.
the exponential model!)
• Musa‐Okumoto Logarithmic Poisson Execution Time
M d l
Model
– Like NHPP, but later fixes have less effect on reliability
• The Delayed S and Inflection S Models
The Delayed S and Inflection S Models
– Delayed S: Recognizes time between failure detection and
fix
– Inflection S: As failures are detected, they reveal more
failures
GRAA
• This model is concerned with modelling the
number of failures observed in given testing
intervals. (1979)
• They proposed that the time‐dependent failure rate
follows an exponential distribution.
e ode s,
• The model is,
[m(t )] y − m (t )
P{N(t)=y}= e , y = 0,1,2...
y!
GRAA