Enhanced and Integrated Ant Colony-Artificial Bee Colony-Based QOS Constrained Multicast Routing for Vanets

Enhanced and Integrated Ant Colony-
Artificial Bee Colony-Based QOS Constrained
Multicast Routing for Vanets
A.Malathi1
,
Research Scholar, Department of Computer Science, Pondicherry Engineering College, Puducherry1
malathiprem@hotmail.com
Dr. N. Sreenath2
,
Professor, Department of Computer Science, Pondicherry Engineering College, Puducherry2
nsreenath@pec.edu
Abstract—The facilitation of inter-vehicular communication
by direct means or through fixed infrastructure devices like
roadside units enable Vehicular Ad hoc NETwork (VANET)
to be the predominant environment for the deployment of
Intelligent Transportation Systems (ITS).The vehicular
movements in VANET is governed by the incorporated
mobility model for achieving better data dissemination rate
as their mode of multicast routes-based parallel transmission
remains indispensable for sharing traffic related
information. The QoS parameters of multicast routing is to
be satisfied essentially on par with the base protocols, place
and time of application for compliance with the necessity in
robust data dissemination. Enhanced and Integrated Ant
Colony-Artificial Bee Colony oriented Multicast Routing
(EIAC-ABCMR) is formulated as a multicast tree
determination problem that imposes the satisfaction of
multi-constrained QoS by reducing cost, delay and jitter
with increased bandwidth for improving efficacy in data
transmission. The simulation experiments of EIAC-ABCMR
and its derived inferences confirm its importance in reducing
the number of multicast groups formed during data
communication which is about 36% predominant to the
compared baseline multicast routing techniques. The results
of EIAC-ABCMR also infers a better throughput, reduced
normalized load overhead and energy consumptions on par
with the benchmarked QoS constrained multicast routing
schemes under study.
Keywords-Ant Colony, Artificial Bee colony,Delayed
Convergence, Multicast Tree, Stagnation.
I. INTRODUCTION
A number of innovative technologies and
predominant contributions were developed by researchers
in the field of VANET for facilitating road safety based on
ITS implementation. These ITS’s developed for road
safety depends on the vehicle-oriented mobility
characteristics like traffic regulations and road conditions
that ends up with dynamic changes in the network
topology [1]. The success of each ITS depend on the
extent of its application and suitability attributed towards
the driver’s safety and ambience in the network. The
drivers of the network need to be periodically informed
about the emergency information that pertains to
accidents, collision avoidance, fuel utility services and
vehicle maintenance services for facilitating hassle free
journey [2-4]. But, these real time significant information
has to be distributed to the vehicle drivers with optimized
jitter, cost, delay and bandwidth in order to ensure QoS
during transmission [5]. Multicasting is the optimal option
for distributing emergency information to multiple groups
of vehicles in the network as they possess the potential of
organizing themselves into temporary groups depending
on the kind of services necessitated by them [6].
Moreover, multicast routing is considered as optimal since
they incur reduced cost in data transmission by
duplicating single packet into multiple packets for
International Journal of Computer Science and Information Security (IJCSIS),
Vol. 15, No. 9, September 2017
42 https://sites.google.com/site/ijcsis/
ISSN 1947-5500

enabling transmission to multiple destination nodes [7].
The optimality of multicast routing can be further
improved by formulating multi-objective function that
imposes multiple QoS constraints for achieving efficient
data dissemination in the network. In the recent past,
potential number of multi-objective QoS-based multicast
routing algorithms were propounded among which
Particle Swarm Optimization(PSO), Firefly Algorithms
(FA), Ant Colony Optimization (ACO), Artificial Bee
Colony(ABC) are identified to be predominant[8-9]. In
addition, limitations like stagnation and delayed
convergence corresponding to ACO and ABC algorithms
open the option of integrating and incorporating them into
the solution of multi-constrained QoS based multicast
routing that estimates optimal multicast tree for routing
[10].
EIAC-ABCMR is an integrated and enhanced multi-
constraint imposed QoS multicast routing scheme that
prevents stagnation and delayed convergence during
optimal multicast tree prediction. The merits of ACO and
ABC algorithms are mutually utilized for estimating
optimal multicast tree in order to reduce overhead
incurred during multicast routing. EIAC-ABCMR
facilitates better exploration and exploitation rate in the
search domain by enforcing global and local searching
process based on pheromone-based ants of ACO and the
employee bee, onlooker bee, scout bee phases of ABC.
The predominance of the EIAC-ABCMR is studied using
ns-2.34 for quantifying its performance with regard to
effective indicator metrics like throughput, multicast
group count, energy consumptions and packet loss.
The subsequent sections of the paper are structured as
follows. Section II presents the key characteristics and
pitfalls of the reviewed existing predominant meta-
heuristic techniques that are especially attributed for
improving multicasting in VANET. Section III highlights
the problem formulation of EIAC-ABCMR with the key
steps involved in its implementation and determination of
optimal multicast tree under multicast routing. Section IV
presents the results and inferences of the multi-perspective
simulation study conducted for judging the potential of the
EIAC-ABCMR. Section V concludes the paper with
major key contributions of the proposed scheme.
II RELATED WORK
From the recent past, a considerable number of
researchers have contributed potential number of meta-
heuristic techniques for improving multicast routing in
VANET by satisfying the multi-constraints of QoS. Some
of the meta-heuristic technique inspired optimal multicast
routing schemes propounded in the past few years are
considered for review in order to understand its merits and
limitations for proposing a new predominant meta-
heuristic technique that could aid in optimal multicasting.
In 2011, a tree-based optimization scheme [11]
was propounded for supporting the multicast optimizing
process that initially establishes reliable paths and later
combines them in constructing multicast tree. The
overhead in the construction of optimal multicast is
predominantly reduced based on the application of ACO
for achieving both local and global optimization in a
distributed manner. This ACO inspired scheme uses
orthogonal property for combining the parameters into an
integral unit as it is necessary for improving the quality of
optimization. Then, Bee Life Algorithm Based Multicast
Routing (BLABMR) [12] was contributed for improving
the optimizing process that estimates suitable multicast
tree for deterministic routing. This BLABMR addresses
the issue of QoS satisfaction by introducing local and
global optimization procedure that adheres to the
objective of reducing delay, cost and jitter under tolerable
thresholds. The performance of BLABMR is found to
exhibit excellence by minimizing the energy
consumptions and packet latency to a determinable level.
Further, an integrated ACO and PSO scheme
[13] was proposed for multicast tree optimization under
multicast routing in VANET. This integrated ACO-PSO
scheme initiates the optimizing process based on the
generation of mobile agents that facilitate searching in the
problem domain. In this integrated approach, the local
optimization is achieved by the updation of pheromones
ISSN 1947-5500

that determine the movements of mobile agents and the
global optimization is facilitated through mobile agents’
random interaction. The multicast trees are optimized for
not only satisfying the QoS constraints but also for
reducing the cost in data dissemination. Another multicast
tree optimization technique based on bee’s pattern of
communication was propounded in [14] for estimating
and establishing optimal multicast routes between the
source and destinations of the network. In this bee-life
based communication method, the multicast tree is
determined by minimizing the normalized load overhead
and mean end-to-end latency under potential packet
delivery with bandwidth utilization. This method of bee-
life based multicast tree optimization technique depends
on the spatial zone as the multicast tree is optimized by
interacting with the eligible and selected neighbour’s of
the network for reducing the overhead in communication.
In 2015, an ACO-Based Multicast Routing
(ACOBMR) mechanism [15] was proposed for improving
the efficacy in multicast routing by maintaining a routing
table that pertains to each individual node of the
multicasting group. The routing table used in this
mechanism stores and updates the data related to the
multicast group members depending on the mobile nodes
decision to join or leave the network. In this approach,
the existing paths of the network are checked periodically
when the source node decides to send data packets to the
destinations of the network. The packets transfer process
by the source node is terminated whenever a reliable path
that satisfies QoS factors are not satisfied. Likewise, an
ACO-based clustering algorithm was proposed in [16] for
improving the process of clustering such that minimum
number of clusters are created for transmission. The
nature of ACO is employed for local and global
optimization of the clustered multicast group for meeting
the requirements of QoS in the network to estimate an
optimal multicast tree. The utilization of ACO in this
scheme improves the quality or fitness of the solution and
innovates the potential of representing an exhaustive
solution that depends on the combination of independent
solutions in the problem domain.
In addition, a Binary Code-based Artificial Bee
Colony (BABC) mechanism [17] was proposed for
resolving the issues that emerge during the construction of
the multicast spanning tree. This BABC approach uses a
two element deviation method that aids in maintaining the
consistency of the solutions that are optimally coded using
binary values. The proposed BABC scheme is determined
to tackle the issues that arise due to the congestive
conditions of the network. BABC is found to determine
the multicast spanning tree with a success rate of 92% by
generating a number of sub optimal solutions in each
iteration. BABC is also adaptable when the tree paths of
the network are necessarily re-built under dynamic
network changes. Artificial Bee Colony based Multicast
Routing (ABCMR) was proposed in [18] for determining
multicast tree that aids in effective group communication.
The onlooker, employee and scout bee phases of ABC are
suitably applied in multicast routing to determine the
routes of the multicast network based on the requirements
met in minimizing jitter, latency and cost. The results of
ABCMR infer its potential in ensuring optimal solutions
in each and every generation such that optimal level of
discrimination between solutions can be performed
effectively.
The literature review of the aforementioned
meta-heuristic multicast routing techniques proposed for
VANET present two core limitations such as i) the
stagnation degree of ACO is high during optimizations
and ii) the convergence latency of ACO is impactful
during the optimization process. Thus EIAC-ABCMR is
proposed by integrating ACO and ABC algorithms in
order to identify optimal multicast tree during multicast
routing process in VANETs.
In the next section, the problem formulation of
EIAC-ABCMR with the potential steps incorporated in
determining multicast optimal tree for efficient
multicasting in VANET is presented.
ISSN 1947-5500

III ENHANCED AND INTEGRATED ANT
COLONY-ARTIFICIAL BEE COLONY ORIENTED
MULTICAST ROUTING (EIAC-ABCMR)
The implementation of Enhanced and Integrated
Ant Colony-Artificial Bee Colony oriented Multicast
Routing (EIAC-ABCMR) comprises of three steps that
includes: a) Development of Multi-QoS constraint
imposed Multicast Routing problem, b) Determination of
optimal multicast tree using Integrated Ant Colony-
Artificial Bee Colony based metaheuristic algorithm and
c) Facilitation of multicast routing using determined
EIAC-ABCMR-based multicast tree.
A. Development of Multi-QoS constraint imposed
Multicast Routing problem for implementing EIAC-
ABCMR
In this EIAC-ABCMR-based QoS constrained
Multicast Routing problem, the topology of VANET is
considered as an undirected weighted graph with vertices
and edges representing vehicular nodes and link
maintained between the vehicular nodes respectively. The
communication link between each vehicular node is
quantified using multiple QoS constraints such as delay,
bandwidth, cost and jitter since they have direct impact
towards the multicast routing activity in VANET. EIAC-
ABCMR is formulated as a minimized objective solution
that determines optimal multicast tree
)),(( MRCAST DSTMR using benefits derived from the
integrated Ant Colony-Artificial Bee Colony algorithms
for local and global searching process. This optimal
multicast tree is selected from a number of multicast trees
that could be possibly established by the mobile nodes
that exist between the source and destinations of the
network under minimum delay, cost and jitter with
maximum bandwidth utility. Thus EIAC-ABCMR-based
QoS imposed multicast routing solution is formulated as a
minimizing multi-objective function represented through
Minimise
)()()()())),((( bdccjbdaRCASTobj fQoSwfQoSwfQoSwfQoSwDSTMRMf +++=
(1)
which satisfies the QoS constraints of
TH
T
t
d
DelaytspDelayMaximumfQoS ≤= ∑
=
))),((()(
1
(2)
TH
T
t
j JittertspJitterMaximumfQoS ≤= ∑
=
))),((()(
1
(3)
))),(((cos)( RCASTc
DSTMRtfQoS = (4)
Bandwidth
Tt
b NtspbandwidthimumMinfQoS ≤=
∈
)),((()( (5)
In this context, the path determined between the source
and destinations in time’t ‘ is represented using ‘ ),( tsp ‘
with delay threshold ( THDelay ), jitter threshold (
THJitter ) and necessary bandwidth( BandwidthN )
respectively. The objective weights ‘ aw ‘, ‘ bw ’, ‘ cw ’
and ‘ dw ‘ used in the objective function of EIAC-
ABCMR corresponds to the intensity of each QoS
constraint imposed on multicast routing depending upon
the context of necessity.
B. Determination of optimal multicast tree using
Integrated Ant Colony-Artificial Bee Colony based
metaheuristic algorithm
EIAC-ABCMR is an improved multicast routing
solution of an integrated Ant Colony-Artificial Bee
Colony based metaheuristic algorithm that estimates an
optimal multicast tree by preventing stagnation and
delayed convergence in Ant Colony and Artificial Bee
Colony algorithms respectively. The time incurred by
ABC for identifying initial solutions are completely
eliminated by the ants of ACO and similarly the ants use
the information derived from the employee and onlooker
bee step of ABC for estimating optimal solution which
does not require any further exploitation in determining
solution.
In EIAC-ABCMR, the number of ants, employee
bees and onlooker bees are initially set in proportional to
the number of QoS parameters that are needed to be
explored on the routes of each multicast tree. Then, an
optimal multicast tree is selected from a number of
generated multicast trees of the network based on their
satisfaction of minimum cost, delay, jitter and maximum
ISSN 1947-5500

bandwidth. The pheromone value for each QoS parameter
is also assigned initially in order to facilitate the ant
searching process of ACO. The ant searching process of
ACO is initiated by selecting a subset of QoS parameter
from random collections of probable parameters based on
the estimated probability derived using Equation (6)
)()()( iviviFS
PhPhP Δ∗= (6)
Where )(ivPhΔ corresponds to the collection of ants that
selects a subset of QoS parameter ‘i ‘from a random
collection of QoS parameter with the pheromone value
assigned to ‘ )(ivPh ‘. The pheromone value assigned to
each QoS parameter is updated by ants based on selected
and explored distinct subset of QoS parameters using
Equation (7)
)(0)(
)1( iviv
PhPhPh χχ −+= (7)
Where ‘ 10 ≤≤ χ ‘is the factor of relative significance.
When a single iteration of initial basic solutions are
identified through the ants of ACO, the derived subset of
QoS parameter are considered as input to the employee
phase of ABC algorithm for further exploitation. The time
which is essential for estimating initial basic solutions are
completely reduced by the ants-based searching process of
ACO. The subset of QoS parameters elucidated from
ACO serves as its exploitation search space which is
analogous to the food sources of ABC algorithm as
represented in Equation(8).
N
kKSF SPSFQoS ∈)(,)( (8)
Where ‘ kS ‘is the ‘
th
k ‘subset of QoS parameters passed
as input to ABC algorithm from the complete set of QoS
parameters ( )( KSF ) chosen by ACO with ‘
N
SP ‘, ‘ N ‘
pertaining to the exhaustive problem search space and
used dimensionality necessary for exploiting the search
space respectively. The precision of identifying an optimal
solution (multicast tree) from the available initial basic
solutions depend on the possible number of optimal
solutions that has the eligibility of being selected. The
employee bee phase of ABC calculates the fitness value
for each specific initial basic solutions to determine the
optimal solution through Equation (9) depending on the
eligible number of optimal solutions.
))),(((1
1
RCASTobj
k
DSTMRMf
Fitness
+
= (9)
Where ‘ ))),((( RCASTobj DSTMRMf ‘ is the multi-
constrained objective function of EIAC-ABCMR
presented through Equation (1) which aids in
discriminating the initial basic solutions from one another
depending on the imposed constraints of jitter, delay, cost
and bandwidth. Once the classification between the initial
solutions of the problem is achieved by the employee bee
phase of ABC, the onlooker bee phase uses them for
further exploitation. The onlooker bee phase focusing on
the exploitation of each selected solution compares it with
the other neighbourhood solutions for verifying its
optimality (i.e., the objective solution of each multicast
trees is compared with the other objective solution of
neighbourhood multicast trees) using probability ( FSP )
quantified through Equation (10)
∑
=
= MF
l
l
k
kFS
Fitness
Fitness
P
1
)( (10)
Where ‘ kFitness ‘refers to the onlooker bee selected
neighbourhood solutions’ fitness value. If the fitness
value of the selected neighbourhood solutions is
determined to be greater than the initial optimal solutions
considered for exploitation, the onlooker bee phase
updates the old solution ( )(( kK SFPA ) into new
optimal solution( ))(( kNEW SFS ) based on Equation (11)
)))(()((())(())(( NNkkkkkNEW SFPASFPASFPASFS −+= χ (11)
Where ‘ ))(( kk SFPA ’ and ‘ ))(( NN SFPA ’ corresponds to
the reliable accuracy of the solution which is currently
being exploited and the neighbourhood solution which is
to be exploited next. The parameter ‘ χ ‘is used as the
control factor that aids the selection of reliable
neighbourhood solution that lies around the initial basic
ISSN 1947-5500

solutions in the problem space and it is also responsible
for comparing the features(QoS parameters) of each
solution(multicast tree). In this context, the deviation
between ))(( kk SFPA and ))(( NN SFPA leads to delayed
convergence in determining optimal solutions through
ABC. Thus the step length is reactively decreased when
the process of solution converges to the optimal solution
(multicast tree satisfying multiple constraints of QoS).
This discrimination of determining the best solutions
based on Equation (6) continues until a predefined
generations and the identified subset of optimal solutions
are decided and returned back to the ACO. Then the
process of global searching is again initiated by updating
the pheromone value estimated based on Equation (12)
α
ρρ )
1
()1(
)(
)()(
iBest
iFSiFS
S
PP +−= (12)
Where ‘ ρ ‘ and ‘ )(iBestS ‘refers to the pheromone
evaporation rate and ant derived optimal solution over
predefined generation. The new solutions are again
explored based on the updated pheromone value derived
using Equation (7) for enabling the ants of ACO to
determine the optimal solutions through repeated
exploitation by calculating probability based on Equation
(1). Again, the new initial basic solutions are derived and
returned as input to the ABC for a number of generations
until optimal solutions that portray greatest prediction
accuracy is achieved. The optimal solutions that fail to
satisfy the constraints are again passed to the scout bee
phase of ABC for estimating its possibility if feasible
based on probability derived using Equation(5). This
achievement of maximum prediction accuracy in EIAC-
ABCMR is mainly due to the hybridization of ACO and
ABC meta-heuristic schemes employed for estimating
optimal solutions (multicast tree satisfying multiple QoS
constraints under multicast routing in VANET).
C. Enforcement of Multicast routing using EIAC-
ABCMR-based optimal multicast tree.
The optimal solution obtained through EIAC-
ABCMR highlights a selected multicast tree that contains
distinct paths between the source and destinations. The
intermediate nodes with the source and destination from
each distinct path of the selected multicast tree are
represented using the method of string encoding [19].
Vehicular node’s location, identity and their connectivity
links are represented as topological information to the
search space of VANET. The search space of VANET
used in EIAC-ABCMR also includes potential influencing
QoS constraints like delay, jitter, bandwidth and cost for
estimating link quality existing between the vehicular
nodes. The number of possible multicast trees that could
be generated from the considered search space
corresponds to the number of initial solutions that are
initially explored by the ants of ACO in EIAC-ABCMR
scheme. The method of weight aggregation [20] is utilized
in EIAC-ABCMR as analogous to BLABMR since the
impact of intensification in ABC algorithm can only be
ideally sustained by this technique. The solutions
(multicast tree) derived using EIAC-ABCMR are
repeatedly compared and updated unless the minimizing
QoS constraints used in formulating the multi-objective
functions are satisfied.
In the subsequent section, Algorithm 1 which
highlights the significance of EIAC-ABCMR in enforcing
multiple QoS constraints for multicast routing is
presented.
Algorithm 1- Multiple QoS constrained multicast
routing using EIAC-ABCMR
1. Generation=1
2. Predefine the parameters of ACO
2.1 Assign the number of ants ‘ g ‘ equally
proportional to the feasible number of
solutions ‘ f ’ (multicast trees)
2.2 For each feasible solution from 1 to f ,
assign pheromone value to each solution
( )(ivPh ).
3. Predefine the parameters of ABC
ISSN 1947-5500

3.1 Assign the number of employee bees and
onlooker bees equal to the ant count of
ACO.
4. Ant-based exploration (Iterate for each feasible
solution from 1 to f )
4.1 Compute the probability of selecting a
solution by ant from ‘ f ‘ solutions
4.2 Update the explored solution ( )(ivPh ) based
on predefined pheromone value.
4.3 End until ‘ g ‘ ants have completed their
exploration.
5. Employee bee-based exploration (Iterate for
each feasible solution from 1 to f )
5.1 Compute objective function ( )(kSPF ) based
on estimated fitness value ( kFitness ).
5.2 Select optimal solutions for onlooker bee
phase of ABC.
5.3 Estimate the number of onlooker bees
essential for exploiting the optimal solutions
derived from employee bee phase.
6. Employee bee-based exploitation (Iterate for
each optimal solution from 1 to c )
6.1 Discriminate between the determined
optimal solutions based on classifier that
uses greedy approach.
6.2 Eliminate the optimal solutions that fail to
comply with the imposed multi-constrained
QoS parameters.
6.3 Return back the optimal solution to the ant
phase of ACO.
7. Return to Ant-based exploration
7.1 Globally update the pheromone value
7.2 Update the optimal solution of ACO
8. Until generation=generation+1.
9. Scout bee-based exploration of neglected optimal
solutions.
9.1 The neglected solutions are again explored
based on fitness value ( kFitness )
9.2 Estimate best solution from the neglected
solution if feasible
10. Record optimal solutions and choose best among
them until predefined count of generations for
enabling multicast routing.
In the next section, the importance of EIAC-ABCMR
in multicast routing action of VANET is studied and
presented based on simulation-based experiments with its
derived inferences.
IV SIMULATION RESULTS AND INFERENCES
The predominance of EIAC-ABCMR is
evaluated and compared with the techniques of
ABCBMR, ACOBMR and BLABMR by deploying a
simulated VANET test-bed which is implemented over
ns-2.34 and Linux Ubuntu 10.04. C++ is used for coding
EIAC-ABCMR as it motivates the option of integrating
routing protocol with the proposed scheme as both are
implemented through the same programing language. The
simulated network environment deployed for
implementing EIAC-ABCMR consists of 100 vehicular
nodes that are labeled through numbers 0 to 99 with the
velocity of vehicular movement varying from 5 m/sec to
25 m/sec. In this simulation environment, the node’0’ is
considered to be the source node with nodes ‘3’, ‘13’,
‘23’, ‘33’ ,’44’,’55’,’66’,’77’, ‘88’ and ‘99’ considered as
the multicast group leaders for data communication. The
vehicular nodes of implemented test bed randomly move
around 1200x1200 square meters of terrain area with 200
seconds assigned as the simulation time for each run. DSR
protocol is used as the base routing protocol for
implementing EIAC-ABCMR with 500 meters as the
maximum transmission range. In addition, Two Ray
Ground and Omni-directional antenna is used as the
propagation model and radio antenna type necessary for
implementing EIAC-ABCMR scheme. In this
implementation of EIAC-ABCMR technique, the
parameters such as THDelay , THJitter and BandwidthN
are defined to 1500 msec, 25 msec and 700 kbits per
seconds under the weights of 1, 10,1 and 10 respectively.
The vehicular nodes of the network are responsible for
ISSN 1947-5500

estimating an optimal multicast tree from the group of
nodes in the network for enhancing efficacy in data
routing and hence EIAC-ABCMR technique uses 80
seconds as the period of experimentation.
In the first experiment, the performance of
EIAC-ABCMR is compared with ABCBMR, ACOBMR
and BLABMR based on number of network nodes. Fig. 1
presents the plot of throughput for EIAC-ABCMR,
ABCBMR, ACOBMR and BLABMR evaluated under
different network nodes. The throughput of EIAC-
ABCMR is inferred to be systematically increasing under
increasing network nodes as it embeds the integration of
ACO and ABC for enhancing the probability of data
dissemination. The throughput of EIAC-ABCMR is
realized to get improved by 18%, 15% and 11%
predominant to ABCBMR, ACOBMR and BLABMR
multicasting schemes. Fig.s 2, 3 and 4 demonstrate the
efficacy of EIAC-ABCMR over ABCBMR, ACOBMR
and BLABMR in reducing the normalized load overhead,
number of multicast groups and energy consumptions
with a corresponding increase in network nodes. The
performance of EIAC-ABCMR reveals a better reduction
rate of 13%, 11% and 8% in normalized load overhead
compared to ABCBMR, ACOBMR and BLABMR. In
EIAC-ABCMR, the number of multicast groups formed in
the network is reduced by 21%, 17% and 14% exceptional
to ABCBMR, ACOBMR and BLABMR by integrating
the merits of ACO and ABC algorithms for determining
optimal multicast tree. The energy consumption is also
validated to be minimized by 18%, 15% and 12%
compared to the studied multicasting approaches.
Fig. 1 EIAC-ABCMR-Throughput based on network
nodes
Fig. 2 EIAC-ABCMR -Normalized load
overheadbased on different network nodes
Fig. 3 EIAC-ABCMR -Multicast groups formed under
different network nodes
Fig. 4 EIAC-ABCMR- Energy consumptions under
different network nodes
In the second experimental evaluation, the predominance
of EIAC-ABCMR is compared with ABCBMR,
10 20 30 40 50 60 70 80 90 100
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
x 10
5
NETWORK NODES
THROUGHPUT(bits/sec)
EIAC-ABCMR
ABCBMR
ACOBMR
BLABMR
10 20 30 40 50 60 70 80 90 100
3500
4000
4500
5000
5500
6000
6500
7000
NETWORK NODES
NORMALIZEDLOADOVERHEAD
EIAC-ABCMR
ABCBMR
ACOBMR
BLABMR
10 20 30 40 50 60 70 80 90 100
5
10
15
20
25
30
35
40
45
NETWORK NODES
NUMBEROFMULTICASTGROUPS
EIAC-ABCMR
ABCBMR
ACOBMR
BLABMR
10 20 30 40 50 60 70 80 90 100
50
100
150
200
250
300
350
NETWORK NODES
ENERGYCONSUMPTIONS(mJ)
EIAC-ABCMR
ABCBMR
ACOBMR
BLABMR
ISSN 1947-5500

ACOBMR and BLABMR schemes using packet delivery
ratio, end-to-end delay, number of multicast groups and
packet loss studied under different grid size of the
network. Fig. 5 establishes the performance of EIAC-
ABCMR evaluated in terms of packet delivery ratio under
different grid size. The percentage of packet delivery ratio
is determined to be remarkably enhanced by 17%, 14%
and 11% compared to ABCBMR, ACOBMR and
BLABMR approaches. Similarly, Fig.s 6, 7 and 8 reveals
the reduced rate of end-to-end delay, generated multicast
group count and packet loss studied under different grid
size. The end-to-end delay and the generated multicast
group count in EIAC-ABCMR is revealed to be evidently
reduced by 21%, 16%, 13% and 17%, 14%, 11%
correspondingly predominant to the benchmarked
techniques as it is reduces the degree of stagnation and the
rate of delayed convergence in the integrated ACO-ABC
algorithm. The packet loss of EIAC-ABCMR is also
confirmed to be remarkably minimized by 17%, 14% and
12% on par with the multi-constrained QoS approaches
considered for investigation.
Fig. 5 EIAC-ABCMR-Packet delivery ratio under
different grid size
Fig. 6 EIAC-ABCMR–End-to-end delay under
different grid size
Fig. 7 EIAC-ABCMR -Multicast groups formed under
different gird size
Fig. 8 EIAC-ABCMR- Packet loss under different grid
size
In the third experimental investigation, EIAC-ABCMR is
analyzed based on normalized routing load, control
1000 1500 2000 2500 3000 3500 4000
55
60
65
70
75
80
85
90
95
GRID SIZE(in metres)
PACKETDELIVERYRATIO
EIAC-ABCMR
ABCBMR
ACOBMR
BLABMR
1000 1500 2000 2500 3000 3500 4000
0.3
0.32
0.34
0.36
0.38
0.4
0.42
0.44
0.46
0.48
END-TO-ENDDELAY(msec)
EIAC-ABCMR
ABCBMR
ACOBMR
BLABMR
1000 1500 2000 2500 3000 3500 4000
5
10
15
20
25
30
35
40
45
NUMBEROFMULTICASTGROUPS
EIAC-ABCMR
ABCBMR
ACOBMR
BLABMR
1000 1500 2000 2500 3000 3500 4000
15
20
25
30
35
40
45
50
PACKETLOSS(in%)
EIAC-ABCMR
ABCBMR
ACOBMR
BLABMR
ISSN 1947-5500

overhead, average packet latency and total overhead under
different transmission range of the network. Fig. 9
validates the performance of EIAC-ABCMR in terms of
normalized routing load identified under systematically
increasing transmission range. EIAC-ABCMR technique
seems to improve the percentage of normalized routing
load by 19%, 15% and 13% compared to ABCBMR,
ACOBMR and BLABMR schemes. Fig. 10 and 11
exhibits the control overhead and packet latency rate of
EIAC-ABCMR, ABCBMR, ACOBMR and BLABMR
evaluated under different transmission ranges. The control
overhead of EIAC-ABCMR is determined to be
minimized by 17%, 14% and 12% exceptional to the
benchmarked schemes and the packet latency rate is
proved to be significantly reduced by 19%, 16% and 14%
as this multi-QoS constrained technique is phenomenal in
estimating the appropriate multicast tree for eliminating
the hurdles in reliable packet dissemination. The total
overhead of EIAC-ABCMR is also found to be
remarkably reduced by 21%, 18% and 15% corresponding
to ABCBMR, ACOBMR and BLABMR techniques as it
improves the degree of exploration and exploitation in
search space by combining the potential’s of ACO and
ABC algorithms.
Fig. 9 EIAC-ABCMR- Normalized routing load under
different transmission range
Fig. 10 EIAC-ABCMR- Control overhead under
Fig. 11 EIAC-ABCMR- Average Packet latency under
Fig. 12 EIAC-ABCMR- Total overhead under
50 100 150 200 250 300 350 400 450 500
30
40
50
60
70
80
90
TRANSMISSION RANGE(in metres)
NORMALIZEDROUTINGLOAD(in%)
EIAC-ABCMR
ABCBMR
ACOBMR
BLABMR
50 100 150 200 250 300 350 400 450 500
3000
3500
4000
4500
5000
5500
6000
6500
CONTROLOVERHEAD(inbytes)
EIAC-ABCMR
ABCBMR
ACOBMR
BLABMR
50 100 150 200 250 300 350 400 450 500
0.32
0.34
0.36
0.38
0.4
0.42
0.44
0.46
0.48
0.5
AVERAGEPACKETLATENCY(inmsec)
EIAC-ABCMR
ABCBMR
ACOBMR
BLABMR
50 100 150 200 250 300 350 400 450 500
1
1.2
1.4
1.6
1.8
2
2.2
2.4
TOTALOVERHEAD
EIAC-ABCMR
ABCBMR
ACOBMR
BLABMR
ISSN 1947-5500

Fig. 13 EIAC-ABCMR-Fitness based on Multicast
groups
Fig. 14 EIAC-ABCMR-Optimal fitness solution based
on population
In addition, the exceptional performance of EIAC-
ABCMR is also explored using multicast group based
fitness value and population-based optimal fitness value
under different number of generations. The multicast
group-based fitness value of EIAC-ABCMR (17725.8970)
is confirmed to be achieved in the 12th
generation and
hence its performance is exceptional by 8%, 6% and 5%
compared to benchmarked ABCBMR, ACOBMR and
BLABMR schemes. The population-based optimal fitness
value of EIAC-ABCMR is also proved to be sustainable
with increasing number of generations by exhibiting an
enhanced performance to about 9%, 8% and 6% on par
with the baseline techniques considered for the study.
V CONCLUSION
EIAC-ABCMR is propounded and presented as
the predominant minimum QoS constraints satisfying
multicast routing mechanism that utilizes the advantages
of ACO and ABC algorithm for ensuring reliable data
delivery in VANET. EIAC-ABCMR prevents stagnation
and latency in convergence corresponding to ACO and
ABC algorithms as it is proved to be the overhead in
optimal multicast tree determination that aids in optimal
routing. The potential of EIAC-ABCMR is further
identified to be improved over the traditional ACO and
ABC algorithms as it enforces a higher degree of
exploration and exploitation in the search domain with
least cost and time. The number of multicast tree formed
by EIAC-ABCMR seems to be potentially reduced under
routing and this minimization is mainly responsible for its
better fitness value in terms of population and multicast
group count. The results of EIAC-ABCMR infer its
potency in determining the optimal fitness value in the
least number of generations which is exceptionally rapid
to about 32%, 27% and 25% remarkable to the baseline
optimization schemes considered for investigation. The
results of EIAC-ABCMR also reveals its significance in
reducing jitter, cost and delay under multicast routing by
24%, 21% and 19% since it integrates the beneficial
aspects of ACO and ABC in optimal multicast tree
selection.
REFERENCES
[1] El Amine Fekair, M., Lakas, A., & Korichi, A. (2016).
CBQoS-Vanet: Cluster-based artificial bee colony
algorithm for QoS routing protocol in VANET. 2016
International Conference on Selected Topics in Mobile &
Wireless Networking (MoWNeT), 2(1), 34-47.
[2] Jemaa, I. B., Shagdar, O., Martinez, F. J., Garrido, P.,
& Nashashibi, F. (2015). Extended mobility management
and routing protocols for internet-to-VANET
multicasting. 2015 12th Annual IEEE Consumer
0
5000
10000
15000
20000
25000
5 10 15 20 25
FITNESSVALUEBASEDON
MULTICASTGROUPS
NUMBER OF GENERATIONS
EIAC-ABCMR ABCMR
ACOBMR BLABMR
0
20000
40000
60000
5 10 15 20 25
OPTIMALFITNESS
SOLUTION-BASEDON
POPULATION
NUMBER OF GENERATIONS
EIAC-ABCMR ABCBMR
ACOBMR BLABMR
ISSN 1947-5500

Communications and Networking Conference
(CCNC), 2(1), 22-30.
[3] Sebastian, A., Tang, M., Feng, Y., & Looi, M. (2010).
A Multicast Routing Scheme for Efficient Safety Message
Dissemination in VANET. 2010 IEEE Wireless
Communication and Networking Conference, 2(1), 12-20.
[4] Ruichun Gu, Haolei Li, Yuesheng Tan, & Jingyu
Wang. (2010). Research on Ad Hoc network QOS
multicast routing based on ant colony algorithm. The 2nd
International Conference on Information Science and
Engineering, 3(1), 23-34.
[5] Hernafi, Y., Ben Ahmed, M., & Bouhorma, M. (2017).
ACO and PSO Algorithms for Developing a New
Communication Model for VANET Applications in Smart
Cities. Wireless Personal Communications, 2(3), 34-46.
[6] Abdel-Kader, R. F. (2011). Hybrid discrete PSO with
GA operators for efficient QoS-multicast routing. Ain
Shams Engineering Journal, 2(1), 21-31.
[7] Deepalakshmi, P., & Radhakrishnan, S. (2011).
Source-Initiated QoS Multicasting Scheme for MANETs
Using ACO. 2011 International Conference on Process
Automation, Control and Computing, 3(1), 12-23.
[8] Jiang, M., Luo, Y., & Yang, S. (2007). Stochastic
convergence analysis and parameter selection of the
standard particle swarm optimization
algorithm. Information Processing Letters, 102(1), 8-16.
[9] Khushaba, R. N., Al-Ani, A., AlSukker, A., & Al-
Jumaily, A. (2014). A Combined Ant Colony and
Differential Evolution Feature Selection Algorithm. Ant
Colony Optimization and Swarm Intelligence, 2(1), 1-12.
[10] Shunmugapriya, P., & Kanmani, S. (2017). A hybrid
algorithm using ant and bee colony optimization for
feature selection and classification (AC-ABC
Hybrid). Swarm and Evolutionary Computation, 17(2), 1-
16.
[11] Wang, H., Xu, H., Yi, S., & Shi, Z. (2011). A tree-
growth based ant colony algorithm for QoS multicast
routing problem. Expert Systems with Applications, 38(9),
11787-11795.
[12] Bitam, S., & Mellouk, A. (2013).Bee life-based multi
constraints multicast routing optimization for vehicular ad
hoc networks. Journal of Network and Computer
Applications, 36(3), 981-991.
[13] Patel, M. K., Kabat, M. R., & Tripathy, C. R. (2014).
A hybrid ACO/PSO based algorithm for QoS multicast
routing problem. Ain Shams Engineering Journal, 5(1),
113-120.
[14]Bitam, S., Mellouk, A., & Fowler, S. (2014). MQBV:
multicast quality of service swarm bee routing for
vehicular ad hoc networks. Wireless Communications and
Mobile Computing, 15(9), 1391-1404.
[15] Anwar, N., & Deng, H. (2015). Ant Colony
Optimization based multicast routing algorithm for mobile
ad hoc networks. 2015 Advances in Wireless and Optical
Communications (RTUWO), 3(1), 45-56.
[16] Aadil, F., Bajwa, K. B., Khan, S., Chaudary, N. M.,
& Akram, A. (2016). CACONET: Ant Colony
Optimization (ACO) Based Clustering Algorithm for
VANET. Plos One, 11(5), 55-65.
[17] Zhang, X., & Zhang, X. (2017). A binary artificial
bee colony algorithm for constructing spanning trees in
vehicular ad hoc networks. Ad Hoc Networks, 58, 198-
204.
[18] Zhang, X., Zhang, X., &Gu, C. (2017). A micro-
artificial bee colony based multicast routing in vehicular
ad hoc networks. Ad Hoc Networks, 58, 213-221.
[19]Nagib, G., &Ali Abdelaal, W. (2010). Network
routing protocol using Genetic Algorithm. International
Journal of Engineering Trends and Technology, 40(5),
272-280.
[20] Babaeizadeh, S., & Ahmad, R. (2016). Constrained
artificial bee colony algorithm for optimization
problems. omega-elsevier, 45(2), 42-55.
ISSN 1947-5500

Enhanced and Integrated Ant Colony-Artificial Bee Colony-Based QOS Constrained Multicast Routing for Vanets

More Related Content

What's hot (19)

Similar to Enhanced and Integrated Ant Colony-Artificial Bee Colony-Based QOS Constrained Multicast Routing for Vanets (20)

Recently uploaded (20)

Enhanced and Integrated Ant Colony-Artificial Bee Colony-Based QOS Constrained Multicast Routing for Vanets