Route optimization in manets with aco and ga

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
_______________________________________________________________________________________________
Volume: 02 Issue: 11 | Nov-2013, Available @ http://www.ijret.org 587
ROUTE OPTIMIZATION IN MANETS WITH ACO AND GA
Pankaj Vidhate1
, Yogita Wankhade2
1
Lecturer, IT dept., S.G.R.E.F’s C.O.E.M, Ahmdnagar, Maharashtra, India, pankajvidhate@gmail.com
2
Lecturer, IT dept., S.G.R.E.F’s C.O.E.M, Ahmednagar, Maharashtra, India, yogita16.wankhade@gmail.com
Abstract
A mobile ad-hoc network consists of a collection of mobile nodes which can communicate with each other with the help of
wireless links without the help of any pre-existing communication infrastructure. Due to the lack of infrastructure in these type of
networks, nodes itself can act as a routers and relay the packets from source to destination. There are so many routing protocols
used in MANET which also maintains connectivity from source to destination when links on these paths are break due to some
causes like node movements, radio propagation, drainage of battery, and wireless interference. Routing is one of the important
issues which are having a significant impact on network performance. Different measures which are concerned with the Quality
of Service are like end to end delay, packet delivery ratio, control overhead, pause time, routing overhead, and distance in
between source and destination pair. Different optimization techniques can be used to find out an available optimal path from
source to destination. In this paper, we are using Ant Colony Optimization for finding out best possible paths, along with Genetic
Algorithm which helps in giving the globally optimal solution from all the best possible paths which were produced by Ant colony
optimization. This proposed algorithm called as GA-API overcomes the delay in packet delivery by producing the shortest path
and also overcomes the problem of communication interruption due to node or link failure by finding multiple paths between pair
of source and destination nodes.
Index Terms: MANET, Genetic Algorithm, Ant colony Optimization, API.
--------------------------------------------------------------------***----------------------------------------------------------------------
1. INTRODUCTION
A mobile ad-hoc network consists of a collection of mobile
nodes which can communicate with each other with the help
of wireless links without the help of any pre-existing
communication infrastructure. Nodes within each other’s
radio range communicate directly via wireless links, while
those that are far apart rely on intermediate nodes to forward
their messages. Mobile ad hoc networks are the self-
organizing and self-configuring wireless networks
consisting of multiple numbers of hops. MANET is made up
of a set of mobile hosts which can freely movable and helps
in delivering or relaying packets on behalf of one another.
Due to the lack of infrastructure in these type of networks,
nodes itself can act as a routers and relay the packets from
source to destination. For deploying MANETs, there is no
need of infrastructure resulting into the use of MANETs in
the scenarios like crowd control, search and disaster rescue
operations and battlefields.
Routing is one of the important issues which are having a
significant impact on network performance. These routing
protocols are playing a vital role in deciding Quality of
Service. Different measures which are concerned with the
Quality of Service are like end to end delay, packet delivery
ratio, control overhead, pause time, routing overhead, and
distance in between source and destination pair. One can
improve the Quality of Service by reserving the resources
which would be used in the path in between source and
destination pair for packets delivery, by minimizing delay
required by a packet to reach to its destination, by reducing
the probability of link failures or providing an alternate
arrangement of path in case of link or node failure, also by
providing security for the nodes which are being used in the
Mobile ad hoc network formation as well as for the nodes
which newly wants to join to the already existed Mobile ad
hoc network. Quality of Service for the routing protocols
also deals with an optimization of the routing paths.
Different optimization techniques can be used to find out an
available optimal path from source to destination. It is not
necessary that the optimal path should be always the shortest
one, but it should the feasible path which gives the best
possible path from source to destination. In this paper, we
are using Ant Colony Optimization for finding out best
possible paths, along with Genetic Algorithm which helps in
giving the globally optimal solution from all the best
possible paths which were produced by Ant colony
optimization. It means ACO provides best possible solution
and GA prevents the algorithm from going into the local
optima.
2. RELATED WORK
2.1 Introduction
The highly dynamic nature of a mobile ad hoc network
results in frequent and unpredictable changes of network
topology, adding difficulty and complexity to routing among
the mobile nodes. The challenges and complexities, coupled
with the critical importance of routing protocol in
establishing communications among mobile nodes, make the
routing area the most active research area with the MANET
domain.

_______________________________________________________________________________________________
Numerous routing protocols and algorithms have been
proposed. The performance of these protocols has been
studied and compared under different types of networks or
network environments and also under different traffic
conditions. Number of different survey papers and
comparative analyses of different mobile ad hoc routing
protocols have been published.
Routing of the packets in the MANET has several issues to
be cope up with, like host mobility, rate of frequent link
failures, rate of link repairs. The rate of link failures and
repairs may be high when nodes move fast [1].
Busola S. Olagbegi and Natarajan Meghanathan presented a
review on energy efficient and secure multicast routing
protocol for Mobile ad hoc network. A multicast routing
protocol is used to support the transmission of data from a
sender to all the receivers of a multicast group by using the
available bandwidth more efficiently, while topology is
frequently changing. Multicasting is improving the
efficiency of the wireless link when it is sending multiple
copies of messages. Secure multicast routing is also playing
an important role in Mobile ad hoc networks. They have
reviewed different energy efficient routing protocols like
Energy efficient Location Aided Routing (EELAR), Power
Aware Localized Routing (PLR), Minimum Energy Routing
Protocol (MER), etc [2].
Ashwani Kush and Sunil Taneja take an effort to provide
Secured Routing over MANET with Power Management. In
this paper hashing and power estimation is done for route
request, reply and not in route error and route erasure
phases. So that less routing overhead occurs [3].
2.2 Ant Colony Optimization
The Quality of Service of routing protocols can also be
enhanced by minimizing the routing overhead. Different
optimization techniques can be used for that, Ant Colony
Optimization is one of them.
The ACO is based on the behavior of foraging of different
ants. When ants are searching for their food, these ants
wander randomly in any direction and after finding food or
resource they get returned to their nest position or colony.
While wandering and coming back to their colony ants are
laying some chemical substance along the path or route
which is called as pheromone. Many ants can take different
routes for reaching to the same food source. The ants, which
take or use the shortest path, reinforce that path with more
quantity of pheromone which helps other ants to follow that
shortest path. In the mean time, more number of other ants
are getting attracted because of this pheromone trail, which
makes the path even stronger. This nature of ants is helpful
in quickly identifying the shortest path. Ants are nothing but
autonomous agents which are interacting through an indirect
communication which is also known as stigmergy, where
stigmergy is one of the indirect forms of communication in
which these individual ant agents leave signals in the
environment. These signals are then being sensed by some
other agents to decide their own way of behaving.
S. Prasad, Y.P. Singh, and C.S. Rai has proposed
Probabilistic Ant Routing algorithm which is based on Ant
Colony Optimization. The Probabilistic Ant Routing
algorithm is a proactive routing protocol and it uses both
pheromone and heuristic value [4].
Gianni Di Caro, Frederick Ducatelle and Luca Maria
Gambardella have worked on AntHocNet. AntHocNet is a
hybrid multipath algorithm, designed by considering the
principles of Ant Colony Optimization routing. AntHocNet
consists of both proactive and reactive components. It does
not maintain routes to all possible destinations at all times
like the original ACO algorithms for wired networks, but
only sets up paths from source to destination when they are
needed while starting of a data session [5].
2.3 Genetic Algorithm
Genetic Algorithms is branch of different computational
models which gets its inspiration from the human evolution.
The solution for a specific problem is being encoded by
these genetic algorithms on a chromosome like data
structure. Then after recombination operators are being
applied to these structures. In this way genetic algorithms
preserve critical information. Genetic algorithms can be
used as a function optimizer, although the ranges of these
problems are very broad to which these algorithms have
been applied. With a random population of different
chromosomes an implementation of genetic algorithm is
begun. After this evaluation of these structures and the
allocated reproductive opportunities has been allocated to
different chromosomes which are representing good solution
for the defined problem are given more chances for
reproducing themselves as compared to the chromosomes
which are producing poorer solutions. The goodness of a
solution is decided with the help of current available
population.
C. Safak Sahin, Ibrahim Hokelek, and Giorgio Bertoli have
presented a force based genetic algorithm [6]. This
algorithm is for self-spreading mobile nodes uniformly over
a geographical area. In this algorithm, each mobile node
exchanges its genetic information of direction and speed
encoded in its chromosomes with its neighboring nodes.
D. Suresh Kumar, K. Manikandan and M.A. Saleem Durai
have worked on secure and adaptive backup routing
protocol for mobile ad hoc networks [7]. In the cases of
failure of links these discovered paths are used. To improve
quality of service of routing, size of the buffer, end to end
required delay and shortest possible paths are used as
parameters for GA in route discovery.
Damla Turgut, Sajal K. Das, Ramez Elmasri and Begumhan
Turgut have worked on a paper representing the usefulness

_______________________________________________________________________________________________
of the genetic algorithms for increasing the performance of
the clustering algorithms in ad hoc networks [8].
3. PROPOSED SCHEME
In Mobile ad hoc networks due to its dynamic topology
changes, there is always a requirement of route discovery
between the pair of nodes. Due to the mobile nature of the
nodes and the different constraints like node failure, link
failure, energy issues of the nodes plays a vital role in the
quality of service of routing algorithm. Because of these
factors it is always necessity to find out the possibilities of
the further routes present in between the pair of nodes. Ant
colony optimization will be very helpful to optimize the
possible available route in between the pair of source and
destination. The optimization in the route discovery and
route selection is required all the time in mobile ad hoc
networks.
This paper introduces a new algorithm which does the
hybridization in between API algorithm (one of the type of
ant colony optimization) and a real coded genetic algorithm
(RCGA). If API is compared to its ACO surrogates, who
were mainly applied to discrete optimization problems, API
is designed for the continuous optimization problems. The
proposed ant-base algorithm is different from the basic ant
colony optimization in terms of search strategy. In this
proposed ant-base algorithm, the ants have to travel with the
help of their memory which stores visual landmarks which
were encountered along with the familiar routes, rather than
using pheromones as used in ACO algorithms. Mainly, API
focuses at maximizing the prey rather than minimizing the
path. The drawback of API is that it may be ending up
quickly in a solution which is a local minimum. The reason
behind ending up quickly is due to a regular movement of
the nest which is only in the best position found by the ants.
API is used to find out the solution for the continuous
optimization problems although it could stop at local
minima, it is useful to find out the solutions which are close
to the globally optimal.
In mobile ad hoc networks due to its dynamic topology
changes, there is always a requirement of route discovery
between the pair of nodes. Different factors like link failure,
node failure, node battery backup leads to the continuous
changes in the mobile ad hoc network causing to find out all
the possible routes available in between source and
destination pair so that if one route gets failed there would
be the another possible route available. In case of the higher
dimension mobile ad hoc networks, there is a large number
of mobile nodes are available. So number of routes can be
available, out of these available routes, the route which is
optimal one should be get selected for the packet delivery.
So the Genetic algorithm creates a population of all these
routes in between the pair of nodes. From this population of
routes the optimal route is getting selected with the help of
Ant colony optimization.
Thus, ant colony optimization algorithms’ percentage of
best solutions (say K %) are added to the Genetic Algorithm
population pool after which the breeding process is followed
proportional to their fitness. Then after in this ACO search
process another percentage is added which is nothing but a
percentage of the best individuals in the Genetic Algorithm
(say L % of the GA population) which gives “fitness-
proportional pheromone”. As like the percentage of best
individuals, another percentage of worst fitted individuals of
Genetic algorithm (say M %) are used to “evaporate a
constant amount of pheromones” in the Ant Colony
Optimization search. The migration would not be taken
place when either algorithms find an improvement or both
algorithms do no improve after iteration.
This paper proposes a hybrid algorithm which stochastic in
nature, intending to solve global unconstrained continuous
optimization problems. These problems can be of medium
and large dimensions. With the help of a hybridization of the
two classes of evolutionary optimization algorithms an
algorithm is proposed in this paper. The two classes of
evolutionary optimization algorithms are as follows: a
special class of Ant Colony Optimization algorithm for
continuous domain and a Real-Coded Genetic Algorithm.
The proposed algorithm is entitled as GA-API, which keeps
the downhill search ability of API. Meanwhile the local
minimum trapping in the API algorithm is avoided by using
diversity in the solution which is given by Real-Coded
Genetic Algorithm. Due to the use of proposed optimization
tool the computational time is get reduced also increases the
robustness and consistency of high quality approximate
solutions.
4. STUDY AND ANALYSIS OF INPUT AND
OUTPUT TO THE PROJECT
Input to API:
API is one of the surrogates of ant colony optimization
algorithm. The mobile ad hoc network information is given
to this API algorithm. This API algorithm uses this
information of all nodes and the links present in between
these nodes and traces out the feasible paths in between the
source and destination pairs. As the nodes in MANETs are
mobile in nature, they are continuously changing their
positions causing the link breakage and new link formation
all the time. This behavior of mobility changes the
information provided to the API algorithm, and API has to
trace out all the available routing paths for this changing
behavior.
Input to GA:
Genetic algorithm basically works with the large quantity of
input. With this input genetic algorithm forms its own
population which is nothing but the solution space for the
given problem. In this project, the real coded genetic
algorithm is used. The output from the API is provided to
this RCGA, which further creates new routing paths from
these routes. These routing paths will leads to the formation

_______________________________________________________________________________________________
of population. From this population, by using different
genetic operators like crossover, mutation etc, next offspring
are created which are again included in the population. Out
of these solutions present in the solution then the good ones
are forwarded to the API algorithm again, so that API
algorithm will not get stuck to the local optima because of
its local solution space.
Input to GA-API:
GA-API algorithm is a combination of Genetic algorithm
and API algorithm. The output from the API algorithm is
provided to the GA-API algorithm for further improvement.
The solutions are track out by GA-API which was provided
by API algorithm. From these solutions the solution which
one is globally optimal is been selected as the optimized
solution or the optimized path from source to destination.
Output Analysis:
The output given by the GA-API algorithm is an optimized
path for routing in between the pairs of source and
destination. The output is analyzed with the help of different
parameters. The parameters used for analysis are such as
end to end delay required for packet delivery, packet
delivery ratio in between the given source and destination
pair, bandwidth consumed while transferring packets, the
amount of battery gets consumed during the communication.
By using all these measures it get decided that whether the
route selected by GA-API algorithm is the optimum one or
not.
5. CONCLUSIONS
In this paper, a new algorithm, called as GA-API is
proposed which is a combination of Genetic Algorithm
along with API algorithm which is nothing but one of the
surrogate of Ant Colony Optimization algorithm. This
algorithm is proposed to solve global unconstrained
continuous optimization problems. By combining some
unique characteristics of two robust meta-heuristic
algorithms such as RCGA and API, the hybrid meta-
heuristic algorithm is proposed in this paper. GA-API
provides satisfactory or optimum solutions, with much less
computational effort. This algorithm works well for large,
complex problems with greater dimensionality. GA-API
gave near global solutions. But GA-API failed to find the
global optimum solution in some cases like for the
optimization problems which are generally of small
dimensions.
ACKNOWLEDGEMENT
I would like to acknowledge all the great people who
have been assisted me right the way through my research
work. I would like to thank my parents and my friends for
encouragement and suggestions at every time for this
research. The acknowledgement would be unfinished
without talk about of the lucky thing of the almighty.
REFERENCES
[1]. Josh Broch, David Maltz, David Johnson, Yih-Chun Hu,
and Jorjeta Jetcheva, “A performance comparison of multi-
hop wireless ad hoc network routing protocols” in
MobiCom’98, ACM, 1998.
[2]. Busola Olagbegi and Natarajan Meghanathan, “A
Review of the Energy Efficient & Secure Multicast Routing
Protocols For MANETs” in International Journal on
applications of graph theory in Wireless ad hoc networks,
vol. 2, No. 2, June 2010.
[3]. Ashwani Kush and Sunil Taneja, “Secured Routing over
MANET with Power Management” in ACM, 2011.
[4]. S.Prasad, Y.P. Singh, and C.S. Rai, “Swarm based
intelligent routing for MANETs” in International Journal on
Recent Trends in Engg., Vol. 1, No. 1, May 2009.
[5]. Gianni Di Caro, Frederick Ducatelle and Luca Maria
Gambardella, “AntHocNet: An Adaptive Nature-Inspired
Algorithm for Routing in Mobile Ad Hoc Networks” in
IDSIA, 2004.
[6]. C. Safak Sahin, Ibrahim Hokelek, and Giorgio Bertoli,
“Genetic Algorithms for Self-Spreading Nodes in
MANETs” in GECCO’08 ACM 2008.
[7] D. Suresh Kumar, K. Manikandan and M.A. Saleem
Durai, “Secure On-Demand Routing Protocol for MANET
using Genetic Algorithm” in International Journal of
Computer Applications Vol. 19, No. 8, April 2011.
[8]. Damla Turgut, Sajal K. Das, Ramez Elmasri and
Begumhan Turgut, “Optimizing Clustering Algorithm in
Mobile Ad hoc Networks Using Genetic Algorithmic
Approach”,2000.
BIOGRAPHIES
Pankaj Vidhate received the Bachelor’s
degree in Information Technology
engineering from University of Pune, India
in 2009 and currently working for M.E.
degree in Computer Networks Engineering
from University of Pune, India.
He is currently lecturer with the SGREF
COE Ahemednagar, University of Pune.
His research interests include optimization problems in
mobile ad hoc networks
Yogita Wankhade received the Bachelor’s
degree in Information Technology
engineering from University of India in
2009 and currently working for M.E.
degree in Computer Networks Engineering
from University of Pune, India

Route optimization in manets with aco and ga

More Related Content

What's hot (18)

Viewers also liked (17)

Similar to Route optimization in manets with aco and ga (20)

More from eSAT Journals (20)

Recently uploaded (20)

Route optimization in manets with aco and ga