International Journal on Recent and Innovation Trends in Computing and Communication

Volume: 4 Issue: 3

ISSN: 2321-8169
408 - 414

_______________________________________________________________________________________

ACO-GA based routing algorithm for MANET

Miss. Nikita sunil Barve.

Pillais institute of information technology

New panvel, Navi mumbai
Email address: niki.barve@gmail.com

Prof. Manjusha Deshmukh.

Pillais institute of information technology

New panvel, Navi mumbai
Email address: mdeshmukh@mes.ac.in

AbstractMANET is made up of a collection of mobile wireless nodes, each node act as router as there is no central coordinator present in
this network to ensure in network routing.In MANETmultiple paths are available for data transmission, but itis required to choosemost efficient
path between multiple pathsby providing better Quality of Service(QoS).Routing in MANET network is a challenging task, because the topology
of the network is frequently changes. This issue is addressed in this paper.Here we introduce new routing mechanism by using ACO and GA
approach to satisfy quality matrix.By using ant algorithm to explore the network using intelligent ANT packets based on pheromone
concentration. Multiple paths are generated by ant algorithm and best path is chosen for data transmission, if link failure is occurs while data
transmission then genetic algorithm is used for handling link failure by selecting another backup path based on fitness value.
KeywordsACO,GA,pheromone, fitness

__________________________________________________*****_________________________________________________
I.
INTRODUCTION
MANET is made up of collection of mobile wireless
nodes. Each node act as router as there is no central
coordinator present in this network to assist in network
routing. Now a days MANETs are used in many applications
like, personal area networkmilitary battlefields,crisis
management services classrooms, etc.In wireless networks, it
is necessary to create paths from one host to another to
exchange information. Routing is the process of finding an
efficient path to transfer data from source to destination while
maintaining Quality.
Routing is a process of finding the paths between nodes.
There are mainly two types of routing mechanisms static
routing and dynamic routing. In static routing, the routes
between the nodes are precomputed and stored in routing
table. If a topology of network changes, then the path between
nodes may also change, and hence in dynamic routing is used.
In dynamic routing routes are not stored previously, routes are
generated when required.All packets between any two nodes
are sent in the same path. The new routes are generated based
on the factors like link utilization, traffic etc formaximum
performance. Some traditional routing protocols are RIP
(Routing Information Protocol), OSPF(open shortest path
first),BGP(border gatway protocol).this protocols are not
supported for MANET.
MANETs uses protocols which are classified into three
categories are: 1) table driven or proactive protocol (DSDV)
which maintains all information of nodes for fixed interval of
time. 2) On demand or reactive routing protocols (ad hoc on
demand vector routing protocol) which establish routes as per
requirement. 3) Hybrid protocols (ZRP) are the combination of
reactive andproactive routing protocol. Following figure
shows protocol list used by mobile ad-hoc network.

Figure 1:different types of protocol for manet

These all protocols are designed using different algorithms

such as Bellman ford, Djkestra, shortest path algorithm etc.
All of these protocols are works better for specific
applications, but are not suitable for all environments. A
Comparisonof reactiveand proactiveprotocolis givenin table1.
Table 1: comparison of protocols
Proactiveprotocol
Maintain consistent, up-todateroutinginformation
betweenevery pair of nodes
Routeupdatesatfixedintervals
Arouteto eachother nodein adhocnetworkisatalltimes available

Reactiveprotocol
Establisha routetoadestination
whenthereisademandfor it
Thereareno periodic updates
Routeto
allnodesnotavailableallthetime

Many protocols are designed as per applications

requirment. It is essential to design an efficient, scalable and
adaptiveprotocol which can make decisions in serious
situations. The main purposes of designing new protocol
ACO-GA is to find a path from source to destination by
satisfying users requirement by improving QoS and repairing
the path quickly in case of link failure without corrupting the
level of QoS. In this protocol we combine two optimization
methods Ant colony optimization and Genetic algorithm.
II. PROPOSED ALGORITHM
The static routing protocols are not suitableBecause of the
dynamic nature of the MANETs. Hence there is a need to
create new a dynamic routing protocol. The dynamic routing
protocol should be able to provide certain level of Quality of
Service (QoS) as required by the application.MANET has
dynamic topology and supports both single hop and multiple
hops configuration. Many algorithms have proposed for
dynamic network as discuss above. On the other hand many of
Protocols are specified forapplication so there is a need to
design an adaptive protocol. The proposed algorithm ACOGA enabled ant colony and Genetic based multipath
routing.the idea is to uses ant colony optimization to find the
possible paths from any source node to destination node for
the given network topology. Once the set of possible routes
arefound based on the pheromone concentration by the
artificial ants, the resulting set of routes forms the initial
population given for the genetic algorithm phase.Then based
on the fitness function and genetic operations, the set of
408

IJRITCC | March 2016, Available @ http://www.ijritcc.org

______________________________________________________________________________________________________

International Journal on Recent and Innovation Trends in Computing and Communication

Volume: 4 Issue: 3

ISSN: 2321-8169
408 - 414

_______________________________________________________________________________________
optimal pathsare identified from the initial population for the
network for any source-destination pair. The GA cycle is
continued until there are failure in network. Bellow flow
diagram describe whole concept of ACO-GA in short.

Ant colony optimization

Find all possible Path for any source to

destination for the given topology and select
best path

This diagram shows the total amount of the pheromone

level. The large value of pheromone path isconsidered as a
shortest path.
In our work, the ACO is used for finding the optimal path
using the distance value and bandwidth value assign to the
each node. The host of the ant is considering as the source
node and their food is representing as the destination. The
current node is act as an ant in routing process for finding the
next shortest nodes in the manet network.
In general the ACO along with routing table and data
packets, uses two ant packets such as forward and backward
ant. The forward ant is used while searching the food and the
backward ant is used when the ant get back to source.
Following is pseudo code for ACO.

If failure is occurs while sending data

if (forward ant)
{
Get the next node
if (the link is available and no loopcaused) then
{

Calculate pheromone value

Update forward ant with network status(stack)

Broadcast forward ant packet to all neighbor

}
else if (no such link exist)
{
Create backward ant and load contents of forward ant
tobackward ant (queue).
send Backward ant towards source along the
highest value ofpheromone path
}
}

Genetic algorithm

Based on the fitness function choose

new node among all, perform crossover
and mutation and create new path

GA cycle is continued until successful

data transmission

III. ANT COLONY OPTIMIZATION

Ant colony optimization is a meta-heuristic technique that
uses behaviour of artificial ants to find solutions to
combinatorialoptimization problems. ACO is based on the
behaviour ofreal ants and enhanced abilities such as memoryof
past actions and knowledge about the distance to
otherlocations. In nature, an individual ant is unable
tocommunicate or effectively hunt for food, but as a
group,ants possess the ability to solve complex problems
andsuccessfully find and collect food for their colony.
Antscommunicate using a chemical substance called as
pheromone.As an ant travels, it deposits a constant amount
ofpheromone that other ants can follow. Each ant moves in
arandom fashion, but when an ant meets apheromone trail, it
must decide whether to track it. If itfollows the trail, the ants
own pheromone reinforces theexisting trail, and the increase in
pheromone increases theprobability of the next ant to selecting
the path. Therefore, themore ants that travel on a path, the
more attractive the pathbecomes for subsequent ants.
Additionally, an ant using ashort route to a food source will
return to the nest sooner andtherefore, mark its path twice,
before other ants return. Thisdirectly influences the selection
probability for the next antleaving the nest.

In ant colony optimization eachnode in the network

periodically launches forward ant towards destination.
Forward ants broadcasted in network for finding destination. It
explores the network to reach to the destination. While
exploring network Forward ant collects the information of
network like time stamp, node identification number, at which
the specific node was visited etc. This information is stored in
the stack of forward ant. When forward ant visiting a next
node it deposits some pheromone on the link to indicate that
the path taken. Forward ant on successfully reaching
destination, gets converted to backward ant. The parameters of
forward ant is copied to the queue of backward ant. Backward
ant backtrack the same path as forward ant followed. It
updates the routing table at each node, based on the
information collected by forward ant . Backward having
higher priority than forward ant. Backward ants are killed once
they reach to the source.

Strengths of ant colony optimization algorithm

Figure 2: ACO behaviour

Providing traffic-adaptive and multipath routing.

Relying on both passive and active information
monitoring and gathering.
Making use of stochastic components.
Not allowing local estimates to have global impact.
Setting up paths in a less selfish way than in pure shortest
path schemes favoring load balancing.
Showing limited sensitivity to parameter settings
Robust to agent failures, provide multipath routing
Take care of data load spreading

IV. GENETIC ALGORITHMS

Genetic Algorithm (GA) is a optimization technique that
describes the biological evolution as the problem solving
409

IJRITCC | March 2016, Available @ http://www.ijritcc.org

______________________________________________________________________________________________________

International Journal on Recent and Innovation Trends in Computing and Communication

Volume: 4 Issue: 3

ISSN: 2321-8169
408 - 414

_______________________________________________________________________________________
technique. GA works on the search space which called
aspopulation. Each element in the population is called as
chromosome. GA first starts with randomly selecting set of
feasible solution from population. Each chromosome is a
solution by itself. Each chromosome is estimated for fitness
and this fitness describe the quality of solution. GA uses
adaptive heuristic searching technique through which finds the
set of best outcome from the existing population. New
offsprings means childs are generated from the selected
chromosomes using operators like selection, crossover and
mutation. A Most fit chromosomes are moved for next
generation. The weaker chromosomes get less chance for
moving to next the generation. A population with average
fitness are increases at each iteration, so by repeating the
process of GA for many iterations, better results are revealed.
GA has been widely studied and inspected on manyfields
of engineering. GA offersanother methods forresolving
problems which cant possible to solve using traditional
methods. GA can be applied for nonlinear programming like
minimum spanning tree, traveling salesman problem,
scheduling problem and etc.

Genetic algorithm has ability to solve problems with

no prior knowledge
The performance of GA is based on efficientrepresentation,
evaluation of fitness function and otherparameters like size of
population, rate of crossover andmutation and the strength of
selection.
V. PROBLEM DEFINITION
Here we combined two optimization methods ACO and
GA for better performance and quality of service. The network
is represented as connected graph G = (V,E), with N nodes.
The goal is to find the path with minimum total cost between
source node Vs and destination Vd, where Vs and V belong to
V. This paper presents the efficient on-demand routing
algorithm using ant colony optimization and genetic
algorithm. It is implemented as two modules are ant algorithm
used forNetwork generation or initialization of routing tables
and optimal path are generated by using ant algorithm. Finally
data is sent along the generated path.
A. Initialization of routing table
Every node of network has a routing table. The table
creates betweenneighbor nodes and nodes (destinations) of
network. This table gives probability to reach particular
destination through particular neighbor. The probability
depends on pheromone value of that link of particular
destination. The table also called pheromone table. The
probability is calculated according to this pheromone value for
that link and optimization constraints.
B. Route request

Figure 3: genetic flowchart

The optima path which is found by ACO algorithm, is

given as a input to the GA. The initialise population contain
the path input which is going to evaluate a fitness for each
path. GA consists of three process
Pseudo code for GA:
Choose initial population(Random)
Repeat (until terminated)
Evaluate each individual Fitness
Prune population(Typically all; If not then the worst)
Select pairs to mate from best-ranked individuals

when any node requires to send data to any destination, it

first looks inthe routing table. If there is any route available for
that destination then source node send data through that route.
Availability of route check as, if any neighbor node available
for that destination, and pheromone value is more than
threshold pheromone value than route of that destination is
available. If route is not available than route request process
get start.
Forward ants are broadcast to search the destination.
Forward ants move in the networknode-to-node and collect
information about the nodes they have visited.When forward
ant reaches to destination it is converted into backward ant.
The local pheromone calculation can be done by using
equation 1.
() = ( + ()

eqn (1)

Replenish population(Selected pairs)

Apply Crossover operator
Apply mutation operator
Check for termination criteria
Loop if not terminating(Repeat from step 2)

Strengths of Genetic algorithm

They are parallel in nature.
They explore solution space in Multiple directions at once.
well suitable for solving problems where the solution space
is massive and time taken to search comprehensively is
very high.
They perform well in problems with complex fitness. If the
function is complex, discontinuous, noisy and changes
over time or has many local optima, then GA gives better
results

Let assume source node A wants to send data to a

destination D without any data loss and delay. First node A
broadcasts the Hello message to all the nodes by sending
ANTHELLOPACKET. By means of receivingHELLO
message neighbouring nodes are discovered.Then source node
A sends a Routing request in form of Forward Ant packet to
destination D through all its neighbours and simultaneously
depositssome pheromone and reinforce it. While roaming to
the destination the Forward Ant upgrades routing table which
includes the information about source address, destination
addressand pheromone value. When the Forward Ant reaches
the destination, it will be converted intoBackward Ant and
sends towards the original source in the form of
acknowledgement. The Backward Antwill take the same path
of the corresponding Forward Ant but in reverse direction.
finally the path selected whose pheromone value is high. As
410

IJRITCC | March 2016, Available @ http://www.ijritcc.org

______________________________________________________________________________________________________

International Journal on Recent and Innovation Trends in Computing and Communication

Volume: 4 Issue: 3

ISSN: 2321-8169
408 - 414

_______________________________________________________________________________________
multiple paths are available, in case of link failures the genetic
approach is used and new path is chosen for uninterrupted
transmission.

}
else
{

Forward backward ant on to link available on queue

Update routing table

}
}
if next node is not available
Kill backward ant
}

C. Data transmission
Figure 4: route discovery phase of forward ant

Once number of path created by ACO, a best path i.e a

path which having highest pheromone value will be selected as
primary path for the data transmission. then data transmission
is carried out with primary path.
D. Route mintainance

Figure 5: route discovery phase of backward ant

When BANT reaches the source, it Calculate Global pheromone

for all set of path ij
() = ( + )

.. eqn

(2)

Calculate Probability for path selection based on global pheromone

value of all the path
=

Processing at each node

On arrival of ant at intermediate node, It is required to find
the type of ant whether it is forward or backword .If forward
ant, it discovers path to destination and if backward ant go
back over path to source by updating routing tables at
intermediate nodes.
Pseudo code for ANT algorithm
if (forward ant)
{
Get the next node
if (the link is available and no loop caused) then
{

Send forward ant to the next node

Calculate pheromone value

Update forward ant with network status (stack)

}
else if (no such link exist)
{

Create backward ant and load information offorward ant to

backward ant (queue).

send Backward ant towards source along thesame path as

forward ant
}
}
if backward ant
{
if current node is source node
{

Store path and kill backward ant

Update routing table

The second phase of the routing algorithm is route

maintenance phase. This phase is responsible for the
improvement of the quality of routes during the
communication. ACO does not need any special packets for
route maintenance. Once the FANT and BANT are established
the pheromone tracks for thesource and destination nodes,
subsequent data packets areused to maintain the path.
E. Failuer handlinling
ACO recognizes a route failure through a missing
acknowledgement. If a certain node gets a ROUTE_ERROR
message for a certain link, it will rst deactivates this link by
setting the pheromone value of that link to 0. Then the node
search for ananother alternative link in its own routing table. If
there exists a alternative link node sends the packet via this
path but can not check forthe quality of the path.if second link
doesnot exist the node informs its neighboursthat they can
handover the packet. Either the packet can be transmitted to
the destination node or the backtrack continues to the source
node. If the packet gets fails to reach the destination, the
source node again initiate a new route discovery phase.This
will be time consuming process to again process route
discovery phase. We use genetic algorithm for failure handling
which removes the process of route discovery phase after route
failure occurs.
F. Genetic algorithm
The design of the GA has phases like genetic
representation,
population
initialization,
fitness
function,selection scheme, crossover and mutation. A routing
path comprises of sequence of nodes in network. The genetic
algorithm is applied to paths which is obtained from the route
discovery that is by using ACO phase.
In this section we determine a backup path from source
node todestination in case of primary path failure. By using
GA we obtain an alternate path to be used in link failure. The
alternate path will be next best path when compared tothe
optimal path. When a primary path get fails wecan recover the
connection by using the backup paths. The process of backup
path
routing
contains
three
main
functions:
connectivitymanagement, path discovery phase and path
maintenance phase.

411
IJRITCC | March 2016, Available @ http://www.ijritcc.org

______________________________________________________________________________________________________

International Journal on Recent and Innovation Trends in Computing and Communication

Volume: 4 Issue: 3

ISSN: 2321-8169
408 - 414

_______________________________________________________________________________________

Initial population

In GA each chromosome represents a multiple paths

which obtained in route discovery.

Fitness function

failure it will send a Route_error message to the sourcenode that that

initiates the routing process. So after getting Route_error message the
source node selects the backup path from where the link has broken
insteadof finding a new route by route discovery process. This is
describe in the Fig.8.

For an obtained solution we should be able to evaluate its quality

accurately which can be done with the help of fitness function. Our
goal of using this GA is to handle link failuer, minimize delay
between source and destination.The fitness of each chromosome
considerd as pheromone value of that node

Selection scheme, crossover and mutation

Selection role improving the quality of population of

chromosomes. The selection of chromosome is done on the result of
fitness function.This section describe different methods for selecting
parents for next generation. Following are some of the selection
methods.
a)

Elitist selection- In this method most fit chromosomes are

sure to be selected for next generation.

b)

Rank selection- In rank selection each individual is

selected based on its rank.

c)

Roulette wheel selection In roulette wheel selection, the

individual is selected based on the relative fitness with its
competitors.

Crossover is done to find the better solution from current one.

every time we choose two chromosomes PATH1 and PATH2 for
crossover. PATH1 and PATH2 should have at least one common
node mentioned as v. Now we have two paths i.e path1(sr) and
path2(sr). Now we have v in both paths can be mentioned as
subpath1(sv)
and
subpath(vr),
subpath2(sv)
and
subpath(vr). Now we exchange the sub path (vr) and (vr). The
population will go through mutation after the crossover had been
performed. Both crossover and mutation may produce infeasible
solution so we check it is acyclic. The crossover can be explained
with the following example.

Figure 8: backup path selection

In above digram S is source node and D is the destination node

and other nodes acts as intermediate nodes. data is been sent initially
in the primary path i.e SAF D .When the node F detects the
link or node failure it sends a Route_ error message to the source
node. Now the source node knows that where the failure has occurred
and it selects the backup path that is from node A it selects E as a
backup path and continues to send the data packet. now there is no
required for reroute discovery mechanism in case of link failure
which avoids timeand power consumption.
Flowchart for ACO-GA

Figure 6: crossover

Figure 7: result of crossover

In the above example A is the common node in both

chromosomes. So by performing GA we obtain new two more
chromosomes in order to obtain optimal value.

Path discovery phase

The backup paths are intersect with the primary paths to establish
a newpath. The primary path and backup path are recognised during
the route discovery phase itself when we discover optimal path by
using aco. The backup paths are geographically closer to the primary
paths.

Path maintenance

The data packets are transported via the primary path till the
primary path is gets failed. When a node notices a link failure, it uses
the backup path instead of primary path. This is done with the help of
Route_error message where, when a node detects the link or node

Figure 9: ACO-GA Flowchart

VI. PERFORMANCE EVALUATION

The performance of ACO-GA protocol is verified by
comparing it with ACO. Network simulator (NS2) is used
simulate network. For given simulation the different
simulation parameters are tabulated in Table 2. Performance
metrics are examined by varying the number mobility and
412

IJRITCC | March 2016, Available @ http://www.ijritcc.org

______________________________________________________________________________________________________

International Journal on Recent and Innovation Trends in Computing and Communication

Volume: 4 Issue: 3

ISSN: 2321-8169
408 - 414

_______________________________________________________________________________________
number of node. the topology is examined for 20,50 and 80
node and it shows that ACO-GA works best in 80
nodes.Results are generated in the form of graph.
The results given below are examined by Miss Nikita
Barve which shows ACO-GA protocol performs better than
ACO protocol.
Table 2: performance table

Parameter

Values

Propagation

TwoRayGround

Simulation time

200s

Simulation area

1000x1000

Antenna
MAC layers protocol
Routing protocol
Traffic pattern
No. of nodes
Max speed

Omini directional
IEEE 802.11
ACO-GA
CBR
20,50,80 nodes
5m/s,10m/s,.25m/s

Figure 11: Packet delivery ratio

Fig 11. It shows the effect of mobility of nodes on packet

delivery ratio. ACO-GA is compare with ACO. As MANET is
set of mobile nodes, these nodes move randomly.in ACO
Increase in speed of nodes leads to decrease in packet delivery
ratio. As the mobility of the nodes increases, the probability of
link failure increases and hence the number of packet drops.
But in ACO-GA packet delivery ratio is increases as speed
increases. ACO-GA has much higher packet delivery ratio
than that of ACO.

Performance metrics are evaluated to check QoS of a

presented protocol are as follows.
1)Packet delivery ratio- It is ratio of number of packets
sent by sources and the number of received packets at the
destination.
2)Throughput- It is the average rate of successful message
delivery over a communication channel. It is measured in bits
per second
3)End to end delay- It is a time required for packets to
reach to destination node from source node.
Now here, results for ACO-GA protocol are given. These
results are obtained for packet delivery ratio, packet loss,
controlled overhead, throughput and average end to end delay
parameters for different speed time.

Figure 12: Packet loss

Fig 12 shows the effect of packet loss .in ACO as speed of

nodes increases packet loss is also increases. In ACO-GA as
speed increases packet loss decreases.packet loss ratio is much
less in ACO-GA as compare to ACO.

Figure 10: throughput

Fig 10 shows throughput of ACO-GA. Which is higher

than ACO

Figure 13: Overhead

Overhead is increases in ACO because of route discovery

phase .in ACO-GA overhead is decreses as mobility of the
nodes increase

413
IJRITCC | March 2016, Available @ http://www.ijritcc.org

______________________________________________________________________________________________________

International Journal on Recent and Innovation Trends in Computing and Communication

Volume: 4 Issue: 3

ISSN: 2321-8169
408 - 414

_______________________________________________________________________________________
[6]

[7]

[8]

[9]

Figure 14: average delay

(IWAHN 2002), Vancouver,British Columbia, Canada,

August 18-21, 2002.
D.Suresh kumar , K.Manikandan , M.A.Saleem
Durai,Secure On-Demand Routing Protocol for MANET
using Genetic Algorithm,International Journal of Computer
Applications (0975 8887) Volume 19 No.8, April 2011.
Cauvery N Kand Dr K V Viswanatha, Routing in Dynamic
Network using Ants and Genetic Algorithm, IJCSNS
International Journal of Computer Science and Network
Security, VOL.9 No.3, March 2009.
S.Nagalakshmi , Dr.Rakesh Poonia , Ant Inspired Adaptive
Routing: MANET Routing Protocols- A Survey,
International Journal of Innovative Research in Science,
Engineering and Technology,2319-875
Said Elaiwat , Ammar Alazab, Sitalakshmi Venkatraman,
MamounAlazab, GOM: New Genetic Optimizing Model
for Broadcasting Treein MANET, 2nd International
Conference on Computer Technology andDevelopment
(ICCTD 2010),ISSN-978-1-4244-8845-2

Fig 14 shows average delay. As compare to other Qos

parameters end to end delay is much high in ACO-GA
protocol, which will be improve in future work
Acknowledgment
It is my pleasure to get this opportunity to thank my
beloved and respected prof. Manjusha deshmukh who
imparted valuable basic knowledge of network specifically
related wireless communication. I am are grateful to
department of Computer engineering and, Pillais institute of
information technology, New panvel for providing us
infrastructure facilities andmoral support.
Conclusion
In this work ant algorithm and genetic algorithm are used for
routing in Manet networks. Ant algorithm is found to reduce
the size of routing table and ti found optimal path through
forward ant and backward ant by using pheromone value. But
it cannot works well in failure handling, thus Genetic
algorithm. Genetic algorithm reduces route discovery phase by
using backup path and also maintain quality of the path.
Hence, the combination of these two algorithms,which makes
the packets to explore the network independently, helps in
finding path between pair of
nodes effectively.
References
[1]

[2]

[3]

[4]

[5]

Rajanigandha Metri, Prof. Sujata Agrawal, Ant Colony

Optimization Algorithm Based An Intelligent Protocol To
Improve QoS of MANETs 2014 International Conference
on Circuits, Systems, Communication and Information
Technology Applications, (CSCITA)978-1-4799-24943/14/$31.00 2014 IEEE.
S.Aravindhand Mr.G.Michael, "HYBRID OF ANT
COLONY
OPTIMIZATION
AND
GENETIC
ALGORITHM FOR SHORTEST PATH IN WIRELESS
MESH NETWORKs", Journal of Global Research in
Computer Science, Volume 3, No. 1, January 2012, ISSN2229-371X
Jaspal Jindal ,Vishal Gupta, "Fuzzy Improved Genetic
Approach for Route Optimization inMANET", International
Journal of Advanced Research in Computer Science and
Software Engineering, Volume 3, Issue 6, June 2013, ISSN:
2277 128X.
Xue-Mei Sun, Xiao-Yu Lv, Novel dynamic ant genetic
algorithmfor QoS routing in wireless mesh networks, 9781-4244-3693-4/09/$25.00 2009 IEEE.
Mesut Gunes, Udo Sorges, Imed Bouazizi,ARA The
Ant-Colony
Based
Routing
Algorithm
for
MANETs,International Workshop on Ad Hoc Networking
414

IJRITCC | March 2016, Available @ http://www.ijritcc.org

______________________________________________________________________________________________________