Professional Documents
Culture Documents
( (
= ( (
(4.3)
1
ij
ij
d
q = (4.4)
(1 )
ij ij
t t = (4.5)
ij
t : Pheromone value for edge connecting city i & j
M : Number of ants
nn
C : Length of path found using a nearest neighbour heuristic
k
ij
p : Probability of ant k selecting the edge connecting city i & j
o : Magnitude of pheromone influence on probabilistic decision
ij
q : Heuristic value for edge connecting city i & j
| : Magnitude of heuristic influence on probabilistic decision
ij
d : The distance between city i & j
: Pheromone evaporation rate
Q: Amount of pheromone to deposit
L : Path length
ij ij
Q
L
t t = +
0
( , ), /
nn
ij
i j m C t t
=
=
M. Tech (ACS), NIT Warangal Page 39
4.2.2 Ant Colony Systems
Ant Colony Systems [37](initially introduced as Ant Q]) differs from AS in three
areas:
1. Introduction of a local pheromone update.
2. Modification of the global pheromone update.
3. Modification of the random proportional rule to become the pseudo-random
proportional rule.
The local pheromone update is applied by all ants during the solution construction phase.
Every ant continually applies the update rule to the last solution component used as in Equ.
3.4.6. The aim of this pheromone update rule is to attempt to diversify the search process as
much as possible during the solution construction phase. Without it most ants will simply
create the same solution which will lead the search into a stagnation behaviour.
0
(1 )
ij ij
t t t = + (4.6)
The global pheromone update is modified so that only the best-so-far or iteration-best
solution updates the pheromone map at the completion of solution construction. This
means that unless a solution component has been included in the best solution it will not
receive any modification from the global pheromone update.
(1 ) if ( , ) .
otherwise,
ij ij
ij
ij
i j belongstothebesttour t t
t
t
+ A e
(4.7)
The value of
IJ
t A reflects the utility of the solution and is dependent on the problem e.g.
for the TSP as in Sec. 4.1 it can simply be the inverse of the path length of the solution. The
final and perhaps most important difference between ACS and AS is the modification of the
random proportional rule to become the pseudo-random proportional rule. This rule
introduces a new parameter q0. When a uniformly random value q in the range [0, 1] is less
than q0, the largest transition probability value generated by Equ.3.4.2 is used, rather than
using a roulette wheel selection of all generated probabilities.
M. Tech (ACS), NIT Warangal Page 40
4.3 Simulation Parameters
One Hundred Wireless Sensor Nodes are deployed randomly in 100m x 100m area each with
initial energy of 1 or 2 Joule. Packet length of 1000 bits is assumed. The energy consumed in
processing of one bit of data both in transmitting and receiving electronics (E
elec
) is taken as
50nJ/bit. The energy consumed in transmitting amplifier (E
amp
) for transmitting a bit for unit
distance is taken as 100pJ/bit/m
2
. The sink or gateway is assumed at the coordinates
(25,150) so that a minimum distance of at least 50m from any node is present.
The fallowing figure shows the random deployment of WSN nodes in 100m X 100m area
Figure 4.2 Random deployment of WSN Nodes
0 10 20 30 40 50 60 70 80 90 100
0
10
20
30
40
50
60
70
80
90
100
100nodes random deployment
length in meters
l
e
n
g
t
h
i
n
m
e
t
e
r
s
M. Tech (ACS), NIT Warangal Page 41
The fallowing figure shows the chain formed in PEGASIS using Ant Colony optimisation
described earlier. The total length of the best chain is 834m.
Figure 4.3: Chain formations in PEGASIS using Ant Colony Optimisation.
0 10 20 30 40 50 60 70 80 90 100
0
10
20
30
40
50
60
70
80
90
100
aco chain formation
20 834.0646
length in meters
l
e
n
g
t
h
i
n
m
e
t
e
r
s
M. Tech (ACS), NIT Warangal Page 42
The fallowing shows the average length of all the ants vs. the best ant tour length. It can be
easily observed that the chain length falls drastically after very few iterations and becomes
almost constant after initial iterations. This testifies that Ant Colony Optimisation gives very
good performance in NP hard problem optimisation.
Figure 4.4: The average length vs. Best length of each iteration.
0 2 4 6 8 10 12 14 16 18 20
800
850
900
950
1000
1050
1100
1150
1200
iterations
l
e
n
g
t
h
i
n
m
e
t
e
r
s
lenght of the chain
best route length
averaga lenghth
M. Tech (ACS), NIT Warangal Page 43
4.4 Homogenous PEGASIS with Ant Colony Optimisation
In homogenous network all the nodes Wireless Sensor Network are having same energy of
1Joule each.
4.4.1 Max Energy Node as Cluster Head
In this case the node having the maximum energy is selected as the cluster head. For this all
the nodes while transmitting the data to the chain leader, also indicate their current energy
and expected next state energy after the present transmission is also inserted. Thus
increasing the packet length and unnecessary, which directly drains the battery of both
transmitting and receiving nodes. Also chain leader has to transmit this high data length
packet to the base station which consumes a lot of energy. To overcome the disadvantages
of this cluster head selection in original Pegasis a new cluster head selection criteria is
proposed.
Figure 4.5: Lifetime of Max Energy PEGASIS using ACO.
0 10 20 30 40 50 60 70 80 90 100
0
1000
2000
3000
4000
5000
6000
100Node Network LifeTime Measurement aco homo max smooth
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
M. Tech (ACS), NIT Warangal Page 44
4.4.2 Cluster Head selected sequentially
In this case the present chain leader selects the next chain leader by just passing to the
immediate neighbour in the chain. This mechanism greatly eliminates the need for larger
packet size and conserves a lot of energy in both transmission and reception electronics.
The figure 3.7 shows the lifetime of the Wireless sensor network in this scenario.
Figure4.6: Lifetime of Sequential PEGASIS using ACO.
4.4.3 Comparison of Max Energy and Sequential cluster Head scenarios
From the above figures it can be concluded that the WSN with cluster head selected
sequentially has more lifetime than the one having the max energy node as cluster head.
10% of nodes are dead at around 5300 rounds in former case and around 7000 rounds in
the latter case, an improvement of nearly 32% of the network lifetime. Similarly for 50% of
dead nodes case it is 6800 for former case and it is 8200 for later case. For 100% dead node
case it is 6800 and 9100 round respectively. This can be summarised in a table as fallows.
0 10 20 30 40 50 60 70 80 90 100
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
100Node Network LifeTime Measurement aco hom seq smooth
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
M. Tech (ACS), NIT Warangal Page 45
ACO HOMO Max Energy Sequential % improvement
10% 5300 7000 32
50% 6800 8200 20.6
100% 6800 9100 33.8
Table 4.1 Lifetime comparison of Max Energy and Sequential for Homogenous WSN
Figure 4.7 Lifetime comparison of ACO max energy and sequential cluster head
Hence selecting the cluster head sequentially greatly enhances the lifetime of the network,
although few nodes which are far away from base station die sooner than others in this
case.
0 10 20 30 40 50 60 70 80 90 100
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
100comparision of ACO hom max and sequence smooth
ACO homo max lifetime
ACO homo sequence lifetime
M. Tech (ACS), NIT Warangal Page 46
4.5 Heterogeneous PEGASIS with Ant Colony Optimisation
In heterogeneous network 80% of the nodes in Wireless Sensor Network are having energy
of 1Joule each and the remaining 20% are having 2Joule of Energy.
4.5.1 Max Energy Node as Cluster Head
In this case since 20% of nodes are having more energy, only these nodes will become
cluster head more often in the beginning of the network, soon they deplete their energy in
transmission to the base station.
Figure 4.8 Lifetime of ACO Heterogeneous Max Energy
Thus from the figure it can be observed that nodes having more energy are alive till the end
of the lifetime of the network.
0 10 20 30 40 50 60 70 80 90 100
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
100Node Network LifeTime Measurement aco hetero max smooth
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
M. Tech (ACS), NIT Warangal Page 47
4.5.2 Cluster Head selected sequentially
In this case although some nodes are having higher energy, all nodes become cluster head
equally. So the nodes having high energy tend to stay alive till the end of network lifetime
and also the nodes which are far away from base station die soon. Hence network operation
is not feasible after 50% of nodes die.
Figure 4.9 Lifetime of ACO Heterogeneous Sequential cluster head
4.5.3 Comparison of Hetero Max and Sequential scenarios
It can be observed from the above figures that 10% of node die at around 5100 round for
Max Energy where as it is 7200 round for Sequential cluster head selection. 50 % of nodes
die in Max Energy case at around 6400 and it is 8400 for sequential cluster head selection.
Max energy network Is completely down at 8000 rounds whereas it is 15400 for Sequential.
0 10 20 30 40 50 60 70 80 90 100
0
2000
4000
6000
8000
10000
12000
14000
16000
100Node Network LifeTime Measurement aco hetero seq smooth
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
M. Tech (ACS), NIT Warangal Page 48
ACO HETERO Max Energy Sequential % improvement
10% 5100 7200 41.2
50% 6400 8400 31.25
100% 8000 15400 92.5
Table 4.2 Lifetime comparison of Max Energy and Sequential for Heterogeneous WSN
Figure 4.10 Lifetime Comparison of Hetero Max and Sequential
4.6 Conclusion:
Hence it can be concluded that for Ant Colony Optimisation for both Homogenous and
Heterogeneous sequential cluster head selection maximises the lifetime of Wireless Sensor
Network lifetime.
0 10 20 30 40 50 60 70 80 90 100
0
5000
10000
15000
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
100comparision of greedy hetero max and sequence smooth
ACO hetero max lifetime
ACO hetero sequence lifetime
M. Tech (ACS), NIT Warangal Page 49
5. PEGASIS USING GENETIC ALGORITHM
5.1 Introduction
Genetic algorithms are an optimization technique based on natural evolution. They include
the survival of the fittest idea into a search algorithm which provides a method of searching
which does not need to explore every possible solution in the feasible region to obtain a
good result. Genetic algorithms are based on the natural process of evolution. In nature, the
fittest individuals are most likely to survive and mate; therefore the next generation should
be fitter and healthier because they were bred from healthy parents. This same idea is
applied to a problem by first guessing solutions and then combining the fittest solutions to
create a new generation of solutions which should be better than the previous generation.
We also include a random mutation element to account for the occasional mishap in
nature
The genetic algorithm process consists of the following steps:
Encoding
Evaluation
Crossover
Mutation
Decoding
A suitable encoding is found for the solution to our problem so that each possible solution
has a unique encoding and the encoding is some form of a string. The initial population is
then selected, usually at random though alternative techniques using heuristics have also
been proposed. The fitness of each individual in the population is then computed; that is,
how well the individual fits the problem and whether it is near the optimum compared to
the other individuals in the population. This fitness is used to find the individuals probability
of crossover. If an individual has a high probability (which indicates that it is significantly
closer to the optimum than the rest of its generation) then it is more likely to be chosen to
crossover. Crossover is where the two individuals are recombined to create new individuals
which are copied into the new generation. Next mutation occurs. Some individuals are
M. Tech (ACS), NIT Warangal Page 50
chosen randomly to be mutated and then a mutation point is randomly chosen. The
character in the corresponding position of the string is changed. Once this is done, a new
generation has been formed and the process is repeated until some stopping criteria has
been reached. At this point the individual which is closest to the optimum is decoded and
the process is complete.
Genetic algorithm can be diagrammatically be represented as fallows
Figure5.1 General Scheme of Genetic Algorithm
M. Tech (ACS), NIT Warangal Page 51
The pseudo code for Genetic algorithm is presented in the table below.
Table 5.1: The General Pseudo-code for a Genetic Algorithm
5.2 Basic Explanation
Genetic algorithms range from being very straightforward to being quite difficult to
understand. Before proceeding, a basic explanation is required to understand how genetic
algorithms work. We will use the following problem throughout this section. We want to
maximize the function f = 2x2 + 4x 5 over the integers in the set {0, 1, . . . , 15}. By
calculus or brute force we see that f is maximized when x = 1.
5.2.1 Encoding
The encoding process is often the most difficult aspect of solving a problem using genetic
algorithms. When applying them to a specific problem it is often hard to find an appropriate
representation of the solution that will be easy to use in the crossover process. Remember
Procedure genetic algorithm
Begin
Initilize population with randomly generated candidate solutions;
Evaluate each candidate solution;
While (TERMNINATION CONDITION not satisfied) do
Select parents
Crossover pairs of parents to create a offspring
Mutate the offspring
Evaluate the new candidate
Replace the new candidate generating a new population
End while
End;
M. Tech (ACS), NIT Warangal Page 52
that we need to encode many possible solutions to create a population. The traditional way
to represent a solution is with a string of zeros and ones. However genetic algorithms are
not restricted to this encoding. For now we will use a binary string representation.
Consider the problem defined above. Our possible solutions are obviously just numbers, so
our representation is simply the binary form of each number. For instance, the binary
representations of 12 and 7 are 1100 and 0111 respectively. Note that we added a zero to
the beginning of the string 0111 even though it has no real meaning. We did this so that all
the numbers in the set {0, . . . 15} have the same length. These strings are called
chromosomes and each element (or bit) of the string is called a gene.
We now randomly generate many chromosomes and together they are called the
population.
5.2.2 Evaluation
The evaluation function plays an important role in genetic algorithms. We use the
evaluation function to decide how good a chromosome is. The evaluation function usually
comes straight from the problem. In our case the evaluation function would simply be the
function f = 2x2 + 4x 5, and because we are trying to maximize the function, the larger
the value for f, the better. So, in our case, we would evaluate the function with the two
values 7 and 12.
f (7) = 71
f (12) = 241
Obviously 7 is a better solution than 12, and would therefore have a higher fitness. This
fitness is then used to decide the probability that a particular chromosome would be chosen
to contribute to the next generation. We would normalize the scores that we found and
then create a cumulative probability distribution. This is then used in the crossover process.
M. Tech (ACS), NIT Warangal Page 53
The stopping criteria are used in the evaluation process to determine whether or not the
current generation and the best solution found so far are close to the global optimum.
Various stopping criteria can be used, and usually more than one is employed to account for
different possibilities during the running of the program: the optimal solution is found, the
optimal solution is not found, a local optimum is found, etc. The standard stopping criteria
that is used stops the procedure after a given number of iterations. This is so that if we do
not find a local optimum or a global optimum and do not converge to any one point, the
procedure will still stop at some given time. Another stopping criterion, is to stop after the
best solution has not changed over a specified number of iterations. This will usually
happen when we have found an optimum - either local or global - or a point near the
optimum. Another stopping criteria is when the average fitness of the generation is the
same or close to the fitness of the best solution.
5.2.3 Crossover
Crossover can be a fairly straightforward procedure. In our example, which uses the
simplest case of crossover, we randomly choose two chromosomes to crossover, randomly
pick a crossover point, and then switch all genes after that point. For example, using our
chromosomes
V
1
= 0111
v2 = 1100
we could randomly choose the crossover point after the second gene
V
1
= 01 | 11
V
2
= 11 | 00
Switching the genes after the crossover point would give
V
1
= 0100 = 4
V
2
= 1111 = 15
M. Tech (ACS), NIT Warangal Page 54
We now have two new chromosomes which would be moved into the next population,
called the next generation.
Not every chromosome is used in crossover. The evaluation function gives each
chromosome a score which is used to decide that chromosomes probability of crossover.
The chromosomes are chosen to crossover randomly and the chromosomes with the
highest scores are more likely to be chosen. We use the cumulative distribution created in
the evaluation stage to choose the chromosomes. We generate a random number between
zero and one and then choose which chromosome this corresponds to in our distribution.
We do this again to get a pair, then the crossover is performed and both new chromosomes
are moved into the new generation. This will hopefully mean that the next generation will
be better than the last - because only the best chromosomes from the previous generation
were used to create this generation. Crossover continues until the new generation is full.
It is possible to check each new chromosome to make sure it does not already exist in the
new generation. This means that we will get a variety of possible solutions in each
generation, but also that once we have found the optimal solution in one chromosome, the
other chromosomes will probably not be optimal. That means that the average fitness of the
generation can never be as good as the fitness of the optimal chromosome, which could
make deciding when to stop difficult.
It is also possible to move the best solution from the previous generation directly into the
new generation. This means that the best solution can never get any worse since even if on
average the generation is worse, it will still include the best solution so far.
We can also have two point crossovers. In this case we randomly choose two crossover
points and switch the genes between the two points. In our problem we could pick the
points after the first gene and after the third gene.
V
1
= 0 | 11 | 1
M. Tech (ACS), NIT Warangal Page 55
V
2
= 1 | 10 | 0
to get
V
1
= 0101 = 5
V
2
= 1110 = 14
There are many different crossover routines. We often need to change the crossover
routine to make sure that we do not finish with an illegal chromosome - that is, an infeasible
solution. In this way, crossover is very problem specific.
5.2.4 Mutation
Mutation is used so that we do not get trapped in a local optimum. Due to the randomness
of the process we will occasionally have chromosomes near a local optimum but none near
the global optimum. Therefore the chromosomes near the local optimum will be chosen to
crossover because they will have the better fitness and there will be very little chance of
finding the global optimum. So mutation is a completely random way of getting to possible
solutions that would otherwise not be found.
Mutation is performed after crossover by randomly choosing a chromosome in the new
generation to mutate. We then randomly choose a point to mutate and switch that point.
For instance, in our example we had
V
1
= 0111
If we chose the mutation point to be gene three, v
1
would become
V
1
=0101
We simply changed the 1 in position three to a 0. If there had been a 0 in position three
then we would have changed it to a 1. This is extremely easy in our example but we do not
M. Tech (ACS), NIT Warangal Page 56
always use a string of zeros and ones as our chromosome. Like crossover, mutation is
designed specifically for the problem that it is being used on.
Inversion is a different form of mutation. It is sometimes used in appropriate cases later.
Here inversion operator on our basic example.
The inversion operator consists of randomly choosing two inversion points in the string and
then inverting the bits between the two points. For example
V
2
= 1100
We could choose the two points after gene one and after gene three.
V
2
= 1 | 10 | 0
Now, since there are only two genes between our inversion points, we then switch these
two genes to give
V
2
=1010
If we had a larger chromosome, say
V
3
= 110100101001111
we could choose the inversion points after the third point and after the eleventh point.
V
3
= 110 | 10010100 | 1111
Now, we start at the ends of the cut region and switch the genes at either end moving in.
So we get
V
3
= 110001010011111
Essentially we are just reversing (or inverting) the order of the genes in between the two
chosen points.
M. Tech (ACS), NIT Warangal Page 57
5.3 Simulation Parameters
One Hundred Wireless Sensor Nodes are deployed randomly in 100m x 100m area each with
initial energy of 1 or 2 Joule. Packet length of 1000 bits is assumed. The energy consumed in
processing of one bit of data both in transmitting and receiving electronics (E
elec
) is taken as
50nJ/bit. The energy consumed in transmitting amplifier (E
amp
) for transmitting a bit for unit
distance is taken as 100pJ/bit/m
2
. The sink or gateway is assumed at the coordinates
(25,150) so that a minimum distance of at least 50m from any node is present.
The fallowing figure shows the random deployment of WSN nodes in 100m X 100m area
Figure 5.2 Random deployment of WSN Nodes
The fallowing figure shows the chain formed in PEGASIS using Ant Colony optimisation
described earlier. The total length of the best chain is 3175 meters.
0 10 20 30 40 50 60 70 80 90 100
0
10
20
30
40
50
60
70
80
90
100
100nodes random deployment
length in meters
l
e
n
g
t
h
i
n
m
e
t
e
r
s
M. Tech (ACS), NIT Warangal Page 58
Figure 5.3: Chain formations in PEGASIS Genetic Algorithm.
The fallowing figure shows the average length of all the ants vs. the best ant tour length. It
can be easily observed that the chain length falls drastically after very few iterations and
becomes almost constant after initial iterations.
0 10 20 30 40 50 60 70 80 90 100
0
10
20
30
40
50
60
70
80
90
100
gentic algorithm chain formation
200 3175.5676
length in meters
l
e
n
g
t
h
i
n
m
e
t
e
r
s
M. Tech (ACS), NIT Warangal Page 59
Figure 5.4: Average length vs. Best length of each iteration.
It can be observed from the figure the length of the chain decreases slowly and there is no
guarantee for the convergence to optimum solution.
0 20 40 60 80 100 120 140 160 180 200
2800
3000
3200
3400
3600
3800
4000
4200
4400
4600
4800
iterations
l
e
n
g
t
h
i
n
m
e
t
e
r
s
lenght of the chain
best route length
average lenghh in the iteration
M. Tech (ACS), NIT Warangal Page 60
5.4 Homogenous PEGASIS with Ant Colony Optimisation
In homogenous network all the nodes Wireless Sensor Network are having same energy of 1
Joule each.
5.4.1 Max Energy Node as Cluster Head
In this case the node having the maximum energy is selected as the cluster head. For this all
the nodes while transmitting the data to the chain leader, also indicate their current energy
and expected next state energy after the present transmission is also inserted. Thus
increasing the packet length and unnecessary, which directly drains the battery of both
transmitting and receiving nodes. Also chain leader has to transmit this high data length
packet to the base station which consumes a lot of energy. To overcome the disadvantages
of this cluster head selection in original Pegasis a new cluster head selection criteria is
proposed.
Figure5.5: Lifetime of Max Energy PEGASIS using Genetic Algorithm.
0 10 20 30 40 50 60 70 80 90 100
0
500
1000
1500
2000
2500
3000
3500
4000
100Node Network LifeTime Measurement gen hom max smooth
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
M. Tech (ACS), NIT Warangal Page 61
5.4.2 Cluster Head selected sequentially
In this case the present chain leader selects the next chain leader by just passing to the
immediate neighbour in the chain. This mechanism greatly eliminates the need for larger
packet size and conserves a lot of energy in both transmission and reception electronics.
The figure 3.7 shows the lifetime of the Wireless sensor network in this scenario.
Figure 5.6: Lifetime of Sequential PEGASIS using Genetic Algorithm.
5.4.3 Comparison of Max Energy and Sequential cluster Head scenarios
From the above figures it can be concluded that the WSN with cluster head selected
sequentially has more lifetime than the one having the max energy node as cluster head.
10% of nodes are dead at around 1700 rounds in former case and around 2500 rounds in
the latter case, an improvement of nearly 47% of the network lifetime. Similarly for 50% of
ead nodes case it is 3100 for former case and it is 4300 for later case. For 100% dead
nodecase it is 3850 and 7000 round respectively. This can be summarised in a table as
fallows.
0 10 20 30 40 50 60 70 80 90 100
0
1000
2000
3000
4000
5000
6000
7000
8000
100Node Network LifeTime Measurement gen hom seq smooth
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
M. Tech (ACS), NIT Warangal Page 62
GENETIC HOMO Max Energy Sequential % improvement
10% 1700 2500 47
50% 3100 4300 20.6
100% 3850 7000 81.8
Table 5.1 Lifetime comparison of Max Energy and Sequential for Homogenous WSN
Figure 5.7 Lifetime comparison of GENETIC Max Energy and Sequential cluster head
Hence selecting the cluster head sequentially greatly enhances the lifetime of the network,
although few nodes which are far away from base station die sooner than others in this
case.
0 10 20 30 40 50 60 70 80 90 100
0
1000
2000
3000
4000
5000
6000
7000
8000
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
100comparision of GENETIC hom max and sequence smooth
GENETIC homo max lifetime
GENETIC homo sequence lifetime
M. Tech (ACS), NIT Warangal Page 63
5.5 Heterogeneous PEGASIS with Genetic Algorithm
In heterogeneous network 80% of the nodes in Wireless Sensor Network are having energy
of 1Joule each and the remaining 20% are having 2Joule of Energy.
5.5.1 Max Energy Node as Cluster Head
In this case since 20% of nodes are having more energy, only these nodes will become
cluster head more often in the beginning of the network, soon they deplete their energy in
transmission to the base station.
Figure 5.8 Lifetime of GENETIC Heterogeneous Max Energy
Thus from the figure it can be observed that nodes having more energy are alive till the end
of the lifetime of the network.
0 10 20 30 40 50 60 70 80 90 100
0
1000
2000
3000
4000
5000
6000
7000
8000
100Node Network LifeTime Measurement gen hetero max smooth
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
M. Tech (ACS), NIT Warangal Page 64
5.5.2 Cluster Head selected sequentially
In this case although some nodes are having higher energy, all nodes become cluster head
equally. So the nodes having high energy tend to stay alive till the end of network lifetime
and also the nodes which are far away from base station die soon. Hence network operation
is not feasible after 50% of nodes die.
Figure 5.9 Lifetime of GENETIC Heterogeneous Sequential cluster head
5.5.3 Comparison of Hetero Max and Sequential scenarios
It can be observed from the above figures that 10% of node die at around 2500 round for
Max Energy where as it is 2900 round for Sequential cluster head selection. 50% of nodes
die in Max Energy case at around 5200 and it is 5600 for sequential cluster head selection.
Max energy network is completely down at 7800 rounds whereas it is 10300 for Sequential.
0 10 20 30 40 50 60 70 80 90 100
0
2000
4000
6000
8000
10000
12000
100Node Network LifeTime Measurement GENETIC hetero seq smooth
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
M. Tech (ACS), NIT Warangal Page 65
Genetic HETERO Max Energy Sequential % improvement
10% 2500 2900 16
50% 5200 5600 7.7
100% 7800 10300 32
Table 6.2 Lifetime comparison of Max Energy and Sequential for Heterogeneous WSN
Figure 5.10 Lifetime Comparison of Hetero Max and Sequential
5.6 Conclusion:
Hence it can be concluded that for Genetic Algorithm for both Homogenous and
Heterogeneous sequential cluster head selection maximises the lifetime of Wireless Sensor
Network lifetime. In Genetic there is not much increase in improvement, since the length of
the chain is very high, so the distance between the node is playing dominant role rather
than the packet data size. But Genetic algorithm is very simple to understand and to
formulate the process.
0 10 20 30 40 50 60 70 80 90 100
0
2000
4000
6000
8000
10000
12000
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
100comparision of GENETIC hetero max and sequence smooth
GENETIC hetero max lifetime
GENETIC hetero sequence lifetime
M. Tech (ACS), NIT Warangal Page 66
6. COMAPRATIVE STUDY OF GREEDY, ACO AND GENETIC ALGORITHMS
The chain formed using the Wireless Sensor Node is crucial in deciding the lifetime of the
PEGASIS protocol. The length of the chain formed using Greedy, Ant Colony Optimisation
and Genetic algorithm is compared in the fallowing figure.
Figure 6.1: Chain length comparison of GREEDY, ACO and GENETIC Algorithm.
From the figure it can be seen that the chain formed using Ant Colony Optimisation is less
compared to others.
0 20 40 60 80 100 120 140 160 180 200
500
1000
1500
2000
2500
3000
3500
4000
4500
iteration number
l
e
n
g
t
h
i
n
m
e
t
e
r
s
100comparision of chain length
GREEDY
Ant Colony Optimisation
GENETIC ALGORIYHM
M. Tech (ACS), NIT Warangal Page 67
6.1 Lifetime Comparison of Homogenous WSN
Lifetime comparison of both MAX Energy and Sequential cluster head selection for GREEDY,
ACO and GENETIC Algorithms is done.
6.1.1 Lifetime Comparison of MAX Energy
Figure 6.2 Lifetime Comparison of MAX Energy
From the above figure it can be concluded that the Genetic Algorithm lifetime is less
compared to others mainly because of the chain length is very high compared to other
algorithms. Few nodes in GREEDY algorithm die soon in the beginning itself but for ACO up
to great extent of Network Lifetime hardly any dies.
0 10 20 30 40 50 60 70 80 90 100
0
1000
2000
3000
4000
5000
6000
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
100comparision of HOMO max lifetime
GREEDY HOMO MAX lifetime
ACO HOMO MAX lifetime
GENETIC HOMO max
M. Tech (ACS), NIT Warangal Page 68
6.1.2 Lifetime Comparison of Sequential Energy
Figure 6.3 Lifetime comparison of Sequential
From the above figure it can be concluded that the Genetic Algorithm lifetime is less
compared to others mainly because of the chain length is very high compared to other
algorithms. Few nodes in GREEDY algorithm die soon in the beginning itself but for ACO up
to great extent of Network Lifetime hardly any node dies.
0 10 20 30 40 50 60 70 80 90 100
0
2000
4000
6000
8000
10000
12000
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
100comparision of HOMO SEQUENTIAL lifetime
GREEDY HOMO SEQUENTIAL lifetime
ACO HOMO SEQUENTIAL lifetime
GENETIC HOMO SEQUENTIAL
M. Tech (ACS), NIT Warangal Page 69
6.2 Lifetime Comparison of Heterogeneous WSN
Lifetime comparison of both MAX Energy and Sequential cluster head selection for GREEDY,
ACO and GENETIC Algorithms is done.
6.2.1 Lifetime Comparison of MAX Energy
Figure 6.4 Lifetime Comparison of MAX Energy
From the above figure it can be concluded that the Genetic Algorithm lifetime is less
compared to others mainly because of the chain length is very high compared to other
algorithms. Few nodes in GREEDY algorithm die soon in the beginning itself but for ACO up
to great extent of Network Lifetime hardly any dies
0 10 20 30 40 50 60 70 80 90 100
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
100comparision of hetero max lifetime
GREEDY hetero MAX lifetime
ACO hetero MAX lifetime
GENETIC hetero max
M. Tech (ACS), NIT Warangal Page 70
6.2.2 Lifetime Comparison of Sequential Energy
Figure 6.3 Lifetime comparison of Sequential
From the above figure it can be concluded that the Genetic Algorithm lifetime is less
compared to others mainly because of the chain length is very high compared to other
algorithms. Few nodes in GREEDY algorithm die soon in the beginning itself but for ACO up
to great extent of Network Lifetime hardly any dies
6.3 Conclusion
Chain formed by Ant Colony Optimisation is of least length compared to Greedy and Genetic
Algorithm, because of good global optimisation characteristics of the Ant Colony
Optimisation. The rate of convergence of Ant Colony optimisation is also very fast compared
to other algorithms. The chain is optimised to a great extent within little iteration itself.
The WSN Lifetime is high in case of Sequential cluster head selection since the length of the
data packet is less compared to Max Energy cluster head selection criteria.
0 10 20 30 40 50 60 70 80 90 100
0
5000
10000
15000
number of dead nodes in percentage
n
u
m
b
e
r
o
f
r
o
u
n
d
s
100comparision of HETERO Sequentail lifetime
GREEDY HETERO SEQUENTIAL lifetime
ACO HETERO SEQUENTIAL lifetime
GENETIC HETERO SEQUENTIAL
M. Tech (ACS), NIT Warangal Page 71
7. CONCLUSION
7.1 Introduction
The research carried out for this thesis, investigates energy efficient routing algorithms
related to WSNs. A new cluster head selection criteria and maintenance of priority queue at
each node is proposed. This increases the life of WSNs. Ant Colony Optimisation and Genetic
Algorithms are used in making the chain of PEGASIS. Lifetime of WSN under various
scenarios has been investigated. These chapter summaries the work reported in this thesis,
specifying the limitations of the study and provides some suggestions to future work.
Following this introduction, section 7.2 lists the achievements of the research work. Section
7.3 presents some of the future research area that can be extended to this thesis.
7.2 Contribution of Thesis
The first chapter of the thesis introduced to Wireless Sensor Networks, literature survey and
its architecture. It also provides a brief overview of the thesis. The second chapter discussed
routing algorithms in Wireless Sensor Networks. It presented the literature survey of
PEGASIS routing protocol and of Evolutionary algorithms. The chapter 3 described the
PEGASIS protocol and chain formation using GREEDY algorithm. It also gave radio power
model of WSN and various simulation parameters. The chapter 4 described Ant Colony
Optimisation for minimisation of chain length in PEGASIS, thus contributing to the lifetime
of the WSN. The chapter 5 described Genetic Algorithm. The results of studies have been
presented. The chapter 6 presented the comparative study of GREEDY, ACO and Genetic
algorithms. The chain length and Lifetime of PEGSIS using GREEDY, ACO and Genetic
Algorithm has been compared.
The first contribution of the thesis related to use of sequential cluster head for PEGASIS by
eliminating the overhead, enhances the lifetime of the WSNs. Instead of sending the Energy
status of all the nodes to base station, the next cluster head is selected by the present chain
leader, thus contributing to higher lifetime of Wireless Sensor Network by reduced packet
M. Tech (ACS), NIT Warangal Page 72
size. To validate the algorithm Simulations had been carried out using MATLAB. Simulation
results showed better performance of Sequential as compared to MAX energy in terms of
performance metrics like number of alive nodes and total WSN lifetime.
ACO provides better lifetime for nodes compared to other models. It is also seen that ACO is
able to provide high percentage of nodes live for maximum duration. The chain length
formed ACO is least and converges to the optimum in very little iteration. Hence, it suits
most of application of WSNs which require constant monitoring and sending sensed data
packets to a sink at regular intervals of time.
7.3 Future Directions
To conclude the thesis, the following are some suggestions for the future work which can be
done. In this thesis, ACO and Genetic algorithms have been used. Other bio-inspired
algorithms like Stimulated Annealing, Bacterial Forage optimization, artificial Immune
system (significant time and power consuming) can also be compared to ACO and Genetic
algorithm, but the challenge of reducing computational complexity still remains.
Comparable study of computational complexity of different algorithm need to be analysed.
Secondly, security parameter has not been evaluated in this thesis. So, new security based routing
protocols for Wireless Sensor Networks and its validation can be a field of study. Further, the
proposed protocols have to be dumped into WSN nodes and can be tested in a real time
application.
M. Tech (ACS), NIT Warangal Page 73
8. REFERENCES
1. W. Heinzelman, A. Chandrakasan and H. Balakrishnan, Energy-Efficient Communication
Protocol for Wireless Microsensor Networks, International Conference on System
Sciences, vol. 2, Jan. 2000.
2. W. Heinzelman, A. Chandrakasan and H. Balakrishnan, An Application-Specific Protocol
Architecture for Wireless Microsensor Networks, IEEE Trans. Wireless Comm., vol. 1,
pp. 660-670, Oct. 2002.
3. K. Nakano and S. Olariu, Energy-Efficient Initialization Protocols for Single-Hop Radio
Networks with No Collision Detection, IEEE Trans. Parallel and Distributed Systems, vol.
11, no. 8, Aug.2000.
4. K. Nakano and S. Olariu, Randomized Initialization Protocols for Ad Hoc Networks,
IEEE Trans. Parallel and Distributed Systems, vol. 11, no.7, July 2000.
5. Shio Kumar Singh, M P Singh and D K Singh, Applications Classifications, and Selections
of Energy-efficient Routing protocols for Wireless Sensor Networks, vol.1,pp.085-095,
2010.
6. TANG Yong, ZHOU Ming-tian, and ZHANG Xin, Overview of Routing Protocols in
Wireless Sensor Networks, Journal of Software, vol.17, pp.410-421, March 2006.
7. W. R. Heinzelman, A. Chandrakasan, and H. Balakrishnan, Energy efficient
communication protocol For wireless micro sensor networks, Proceedings of the
Hawaii Conference on System Sciences, 2000, pp.3005-3014.
8. S. Lindsey, and C. Raghavendra, PEGASIS: Power-efficient gathering in sensor
information systems, IEEE Aerospace Conference Proceedings, pp.1125-1130, 2002.
9. T. S. Rappaport, Wireless communications: Principles and Practice, Prentice Hall, July
1996.
10. S Ghiasi, Srivastava A., X. Yang, and M. Sarrafzadeh, Optimal energy aware clustering in
sensor networks, 2002.
11. C. Zhou and B. Krishnamachari Localized topology generation mechanisms for wireless
sensor networks, Global Telecommunications Conference (GLOBECOM), IEEE, vol. 3,
2003.
12. M. Younis K.Akkaya, "A survey on routing protocols for wireless sensor networks," Ad
Hoc Networks, vol. 3, no. 3, pp. 325-349, 2005.
13. Cosmin Cirstea, Energy Efficient Routing Protocols for Wireless Sensor Networks: A
Survey, IEEE International Symposium on Design and Technology in Electronic
Packaging (SIITME), pp.277-282, Oct 2011.
M. Tech (ACS), NIT Warangal Page 74
14. S. Lindsey and C. Raghavendra, PEGASIS: Power-Efficient Gathering in Sensor
Information Systems, IEEE Aerospace Conference, vol. 3, pp. 1125-1130, Mar. 2002.
15. Dali Wei, Yichao Jin, Serdar Vural, Klaus Moessner, Rahim Tafazolli, An Energy-Efficient
Clustering Solution for Wireless Sensor Networks, IEEE Transactions on Wireless
Communications, pp. 3973-3983, 2011.
16. Ossama Younis, Sonia Fahmy, HEED: A Hybrid, Energy-Efficient, Distributed Clustering
Approach for Ad Hoc Sensor Networks, vol. 3, NO. 4, pp.366-379, Oct. 2004,
17. Deng Zhixiang and Qi Bensheng, Three-layered Routing Protocol for WSN Based on
LEACH Algorithm Wireless Mobile and Sensor Networks, (CCWMSN07), pp.72-75,
2007.
18. Indu Shukla, Power Efficient Gathering in Sensor Information System (PEGASIS
Protocol).
19. Jian Wan, Daomin Yuan, Xianghua Xu, A Review of Routing Protocols in Wireless
Sensor Networks, International Conference on Wireless Communications, Networking
and Mobile Computing (WiCOM), pp. 1-4, 2008.
20. Tao Liu and Feng Li, Power-efficient Clustering Routing Protocol Based on Applications
in Wireless Sensor Network, International Conference on Wireless Communications,
Networking and Mobile Computing (WiCom), pp. 1-6, 2009.
21. Zheng Gengsheng, Liu Xiaohua and Hu Xingming, The Research of Clustering Protocol
Based on Chain Routing In WSNs, International Conference on Computational
Intelligence and Industrial Applications (PACIIA), pp.292-295, 2009.
22. Hao Wu, A Chain-based Fast Data Aggregation Algorithm Based On Suppositional Cells
for Wireless Sensor Networks, International Conference on Power Electronics and
Intelligent Transportation System (PEITS), vol. 1, pp. 106-109, 2009.
23. Hyunduk Kim, Boseon Yu, Wonik Choi, Moonju Park and Jinseok Chae, Distance Based
Pre-Cluster Head Selection Scheme for a Chain-Based Protocol, IEEE International
Conference on Computer and Information Technology (CIT), pp. 110-115, 2009.
24. M. Tabibzadeh, M. Sarram and F. Adibnia, Hybrid Routing Protocol for Prolonged
network Lifetime in Large Scale Wireless Sensor Network, International Conference on
Information and Multimedia Technology, pp.179-183, 2009.
25. Wenjing Guo, Wei Zhang and Gang Lu, PEGASIS Protocol in Wireless Sensor Network
Based on an Improved Ant Colony Algorithm, International Workshop on Education
Technology and Computer Science, pp.64-67, 2010.
M. Tech (ACS), NIT Warangal Page 75
26. Young-Long Chen, Jia-Sheng Lin, Yung-Fa Huang, Fu-Kai Cheung and Jen-Yung Lin,
Energy Efficiency of a Chain- Based Scheme with Intra-Grid in Wireless Sensor
Networks, International Symposium on Computer, Communication, Control and
Automation, vol.2, pp.484-487, 2010.
27. Yong chang Yu and Yichang Song, An Energy-Efficient Chain-Based Routing Protocol in
Wireless Sensor Network, International Conference on Computer Application and
System Modeling (ICCASM), vol.11, pp. 486-489, 2010.
28. Feng sen, Qi Bing and Tang Liangrui An Improved Energy-Efficient PEGASIS-Based
Protocol in Wireless Sensor Networks, International Conference on Fuzzy Systems and
Knowledge Discovery (FSKD), 2011.
29. Young-Long Chen, Neng-Chung Wang, Chin-Ling Chen and Yu-Cheng Lin A Coverage
Algorithm to Improve the Performance of PEGASIS in Wireless Sensor Networks,
International. Conference on Software Engineering, Artificial Intelligence, Networking
and Parallel/Distributed Computing, vol. 4, pp. 2230-2233, 2011.
30. S. Lindsey, C. Raghavendra, and K.M. Sivalingam, Data gathering algorithms in sensor
networks using energy metrics, IEEE Trans. Parallel and Distributed Systems, vol. 13,
no. 9, pp. 924-935, Sept. 2002.
31. S. D. Muruganathan, D. C. F Ma., R. I. Bhasin and A. O. Fapojuwo, A Centralized Energy-
Efficient Routing Protocol for Wireless Sensor Networks, IEEE Communications
Magazine, pp. 8-13, March 2005.
32. S. Madden, R. Szewczyk, M. J. Franklin, and D. Culler, Supporting Aggregate Queries
Over Ad-Hoc ireless Sensor Networks, IEEE Workshop on Mobile Computing and
Systems Applications, 2002.
33. K. Kalpakis, K. Dasgupta, and P. Namjoshi, Maximum Lifetime Data Gathering and
Aggregation in wireless Sensor Networks, IEEE Networks Conference, 2002.
34. Wendi Rabiner Heinzelman, Anantha Chandrakasan, and Hari Balakrishnan, Energy
Efficient Communication Protocol for Wireless Micro sensor Networks, Hawaii
International Conference on System Sciences, Jan. 2000.
35. M. Dorigo and G. Di Caro, Ant colony optimization: A new meta-heuristic.
36. A. Colorni, M. Dorigo, V. Maniezzo, Distributed optimization by ant colonies,
European Conference on Artificial Life, pp. 134-142, 1991.
37. Belghachi Mohamed and Feham mohammed, QoS Based on Ant Colony Routing for
Wireless Sensor Networks, International Journal on Computer Science and
Telecommunications, vol. 3, Jan. 2012.
M. Tech (ACS), NIT Warangal Page 76
38. M M chandane, S G Bhirud and S V Bonde, Distributed Energy Aware Routing Protocol
for Wireless Sensor Networks International Journal of Computer Applications, vol. 34,
Nov. 2011.
39. K Saleem , N Fisal, S Hafizah, S Kamilah and R A Rashid, A Self Optimised Multipath
Routing Protocol for Wireless Sensor Networks, International Journal of Recent Trends
in Engineering, vol. 2, Nov. 2009.
M. Tech (ACS), NIT Warangal Page 77