JASC: Journal of Applied Science and Computations ISSN NO: 1076-5131

Enhanced MAC Layer Based On Token-Based Adaptive Scheme

Using TDMA for MANET

K.Thamaraiselvi 1 N.Tamilarasi 2
1,2
Department of Computer Science , Sri Akilandeswari Women’s College – Wandiwash
thamarai1016@gmail.com 1, sjarasi08@gmail.com 2

Abstract – Mobile ad-hoc network (MANET) is an emerging technology that comes under the category of a
wireless network. Even though the network assume that all its mobile nodes are trusted, it is impossible in the
real world as few nodes may be malicious. Therefore, it is essential to put forward a mechanism that can provide
security by selecting an optimal route for data forwarding, using a hybrid optimization algorithm, called M-Lion
Whale, for secure routing. A distributed and adaptive hybrid medium access control (DAH-MAC) scheme
enabled mobile ad hoc network (MANET) in Internet-of-Things (IOT) for supporting voice and data services.
To present communication using workflow technology in MANETs, we present a lightweight service-oriented
architecture (SOA)-based multi-engine architecture for workflow systems in MANETs. The multi-objective
optimization model considers several quality of service (QOS) parameters, namely energy, distance, link lifetime,
delay, and trust. In DAH-MAC, A hybrid super frame structure is designed to accommodate packet
transmissions from a varying number of mobile nodes generating either delay-sensitive voice traffic or best-
effort data traffic. Within each super frame, voice nodes with packets to transmit access the channel in a
contention-free period using distributed time division multiple access (TDMA), while data nodes contended for
channel access in the contention period using truncated carrier sense multiple access with collision avoidance (T-
CSMA/CA). The structure will reduce the delay in data transfer, will increase the data rate of packet transfer and
routing delay will be reduced and data recovery will be possible.

Keywords: MANETs, IoT, T-CSMA/CA, DAH-MAC, M-Lion Whale Algorithm, TDMA

I. Introduction

The Internet-of-Things (IoT) has great potentials to be one of the most promising network infrastructures
towards the next generation wireless network evolution. The IOT framework will interconnect a growing
number of heterogeneous objects, i.e., smart phones, sensors and actuators, autonomous devices, via sensors
suitable wireless technologies for ubiquitous Internet access and pervasive information sharing. Within the
framework, various IoT-oriented intelligent applications can be realized, e.g., disaster monitoring and response,
intelligent control for smart homing, and industrial automation. To support an increasing node number and user
demands, an IoT-enabled mobile ad hoc network (MANET) emerges as an important means to provide
seamless Internet access for the end user. A MANET consists of a group of self-organized nodes,
interconnected for communication in a peer-to-peer manner, without any centralized control. IoT enabled
mobile networks have been envisioned as an important revolution for future generation wireless networks,
which facilitate a growing number of smart heterogeneous objects (e.g., smart sensors, home appliances,
monitoring devices, etc.) being connected via suitable wireless technologies to the Internet, in order to achieve
ubiquitous information access and seamless Communication interaction [4].

Volume 5, Issue 10, October/2018 Page No:1312

JASC: Journal of Applied Science and Computations ISSN NO: 1076-5131

A mobile ad hoc network (MANET) is one usual example of IoT-enabled mobile networks, allowing a group of
smart mobile devices (e.g., smartphones, smart sensor nodes, etc.) interconnected in a peer-to-peer manner and
communicating with no relying on any network infrastructure. MANETs are widely deployed for pervasive IoT-
oriented applications, for example, smart home/office networking [4], emergency communications and crisis
response in disaster areas [5]–[7], and tactical networks for the purpose of command interactions [7]. Device-to-
Device (D2D) communication networks have been recently popularized as one of the typical realizations for
IoT-enabled MANETs, which rely on the ad hoc networking of spatially-distributed smart devices for
information relaying and sharing in some post-disaster areas where the communication infrastructures are
temporarily destroyed [6] [7].

The whale optimization algorithm (WOA) is the optimization algorithm which copies the social behavior of the
humpback whales. The WOA identifies the location of the prey by the random or the best search agent, and the
spiral model simulates the bubble-net hunting scheme of the humpback whale. Humpback whales identify the
location of the prey and encircling them. It begins with the set of random solutions. Since the location of the
optimal solution is not known initially, the WOA assumes that the current solution as the target solution. At
each iteration, the position of the search agent is updated by either randomly chosen agent or the best solution.
This process is repeated until the best solution reaches. Lion algorithm (LA) is another optimization algorithm
which preserves the substantial and genuine performance at the same time it produces the optimal solutions.
The LA is based on the social behaviors of the lion. It determines that the best solution depends on two
behaviors of the lion named as territorial defense and territorial takeover and dissipates the old solutions. In
[25], the basic model of the LA is reconstructed by introducing the fertility evaluation phase. Also, crossover
operation and gender clustering method are changed with subsequent strengthening in the survival fight.

This paper proposes a multi-objective security-based routing scheme in MANET using a hybrid optimization
algorithm. The optimization algorithm used for secured routing is M-LionWhale, which integrates LA into
WOA to provide an optimal solution using a goal programming model. The objectives considered in this
algorithm include several QoS parameters, such as energy, delay, link lifetime, distance, and a security parameter,
called trust. Based on the above objectives, the fitness function is formulated. The steps involved in the
proposed routing algorithm are as follows: (i) measuring QoS metrics, (ii) trust level modeling, (iii) discovering
the k-disjoint path, and (iv) optimal path finding based on the considered multiple objectives. Thus, the
proposed M-LionWhale algorithm provides an optimum route for the transmission with preferable packet
delivery ratio (PDR), throughput, delay, and energy.

The organization of the paper is as follows. Section 2 provides the routing techniques used in the literature that
motivated for the development of the proposed methodology. Section 3 presents the system model and problem
statement along with the multiobjective models. Section 4 deliberates the proposed M-LionWhale algorithm for
secure routing with a suitable block diagram and pseudocode.

Motivation

This section describes the techniques and algorithms used in the literature for MANET routing with security,
which motivates in implementing the proposed security-based routing technique.

Volume 5, Issue 10, October/2018 Page No:1313

JASC: Journal of Applied Science and Computations ISSN NO: 1076-5131

II. RELATED WORKS

The [1] whale optimization algorithm (WOA) is the optimization algorithm which copies the social behavior of
the humpback whales. The WOA identifies the location of the prey by the random or the best search agent, and
the spiral model simulates the bubble-net hunting scheme of the humpback whale. Humpback whales identify
the location of the prey and encircling them. It begins with the set of random solutions. Since the location of the
optimal solution is not known initially, the WOA assumes that the current solution as the target solution. At
each iteration, the position of the search agent is updated by either randomly chosen agent or the best solution.
This process is repeated until the best solution reaches. Lion algorithm (LA) is another optimization algorithm
which preserves the substantial and genuine performance at the same time it produces the optimal solutions.
The LA is based on the social behaviors of the lion. It determines that the best solution depends on two
behaviors of the lion named as territorial defense and territorial takeover and dissipates the old solutions.

The distributed and adaptive hybrid MAC scheme (DAH-MAC) [2], in which distributed TDMA is employed
for voice packet transmissions to guarantee a voice packet loss rate bound and truncated CSMA/CA (T-
CSMA/CA) is used for data nodes to contact the channel. Most of the existing contention-based MAC schemes
evaluate the average access delay for voice traffic in a saturation condition or with a constant arrival rate [7],
which is not the case in reality. We use a more accurate on/off model [8] for voice traffic generation, and exploit
voice traffic multiplexing to improve the voice capacity [8]. Since voice service is real-time, a packet not
transmitted after a delay bound should be dropped at the source, and the voice packet delay has to be evaluated
in a stochastic manner [9] for calculating the packet loss probability. In this way, the delay requirement for voice
traffic can be satisfied probabilistically by guaranteeing the voice packet loss rate.

The contention-based MAC schemes with service differentiation are commonly used for supporting
heterogeneous traffic [10]–[14]. The enhanced distributed channel access (EDCA), standardized in IEEE
802.11e [15], is one typical example, in which delay-sensitive real-time traffic is granted smaller arbitration
interframe space (AIFS) and contention window size to access the channel with a higher probability than non-
realtime traffic [9] [12]. It is demonstrated in [15] that the contention window size differentiation among real-
time and non-realtime traffic is superior over the AIFS differentiation in achieving a smaller access delay for the
real-time service in a traffic saturation condition. To grant voice traffic deterministic channel access priority for
further improving the delay performance, busy-tone base contention protocols are proposed [8] [18], in which
each voice node broadcasts a busy-tone signal, instead of decrementing a back off counter, after an idle the
AIFS duration to prevent the contention intervention from data nodes. Even if the contention separation is
achieved between voice traffic and data traffic in busy-tone based protocols, contention collisions still exist and
accumulate among voice (data) nodes themselves after the voice (data) traffic load becomes relatively high,
making the delay (throughput) performance degraded to an unacceptable level.

III. Proposed Work

Consider a single-channel fully connected MANET [19] [17] [18], where each node can receive packet
transmissions from any other node. The fully connected network scenario can be found in various MANET
applications, including office networking in a building or in a university library where users are restricted to
move in certain geographical areas [20], users within close proximity are networked with ad-hoc mode in a
conference site [8], M2M communications in a residential network for a typical IoT-based smart home
applications where home appliances are normally within the communication range of each other [2].

Volume 5, Issue 10, October/2018 Page No:1314

JASC: Journal of Applied Science and Computations ISSN NO: 1076-5131

The channel is assumed error-free, and packet collision occur when more than one node simultaneously initiate
packet transmission attempts. Without any network infrastructure or centralized controller, nodes exchange
local information with each other and make their transmission decisions in a distributed manner. The network
has two types of nodes, voice nodes, and data nodes, generating delay-sensitive voice traffic and best-effort data
traffic, respectively. Each node is identified by its MAC address and a unique node identifier (ID) that can be
randomly selected and included in each transmitted packet [24]. We use Nv and Nd to indicate the total
numbers of voice and data nodes in the network coverage area, respectively. Nodes are mobile with a fairly low
speed, making Nv and Nd change with time.

For delay-sensitive voice traffic, each packet should be successfully transmitted within a bounded delay to
achieve an acceptable voice communications quality; otherwise, the packet will be dropped. Therefore, as the
main QoS metric for voice traffic, packet loss rate should be carefully controlled under a given threshold,
denoted by PL. The generic on/off characteristic of voice traffic allows traffic multiplexing in transmission.
Every voice source node is represented by an on/off model, which is a two-state Markov process with the on
and off states being the talk spurt and silent periods, respectively. Both periods are independent and
exponentially distributed with respective mean and during a talk spurt, voice packets are generated at a constant
rate, _ packet/s. As for best-effort data traffic, data nodes are expected to exploit limited wireless resources to
achieve as high as possible aggregate throughput. It is assumed that every data node always has packets to
transmit. Nodes in the network are synchronized in time, which can be achieved such as by using the 1PPS
signal with a global positioning system (GPS) receiver [17] [18].

Fig. 1 Super frame Structure

In the network, time is partitioned into superframes of constant duration, denoted by TSF, which is set to have
the same duration as the delay bound of voice traffic. Each superframe is further divided into three periods:
control period (CTP), contention-free period (CFP) and a contention period (CP), the duration of which is
denoted by Tctrl, Tcfp and TCP respectively, as shown in Fig. 1.

The control period consists of Nvm fixed-duration ™ minislots, each with a unique minislot sequence number.
It is to support a different number of voice nodes in the network. Every voice node selects a unique minislot
and broadcasts local information in its selected minislot, for distributed TDMA time slot allocation in the
following contention-free period [32]. In the context of higher service priority to voice traffic, to avoid a
complete deprivation of data service, there is a maximum fraction of time, ‘ (< 1), for voice traffic in each
superframe. The value of ‘is assumed known to all nodes when the network operation starts and can be
broadcast by the existing nodes in each control period.

The voice capacity is the maximum number of voice nodes allowable in the network, denoted by Nvm (same as
the number of minislots in each CTP), under the QoS constraint, which depends on ‘. The period following the
control period is the CFP, which is further divided into several equal-duration TDMA time slots, each slot
having a unique sequence number. Each voice node with packets to transmit (referred to as active voice node)
occupy one-time slot to transmit a number of voice packets, called a voice burst1.

Volume 5, Issue 10, October/2018 Page No:1315

JASC: Journal of Applied Science and Computations ISSN NO: 1076-5131

Thus, the number of TDMA slots in the CFP is determined by the number of voice burst transmissions
scheduled for the superframe, denoted by Ns (Ns _ Nv). The last period CP is devoted to best-effort data nodes
for transmission according to T-CSMA/CA. Data packet transmissions are based on CSMA/CA and are
periodically interrupted by the presence of CTP and CFP.

In the following, we demonstrate how voice nodes access their TDMA slots in each CFP without a central
controller. For data nodes accessing the channel using T-CSMA/CA, we highlight difference between the T-
CSMA/CA within the proposed hybrid superframe structure and the traditional CSMA/CA.

A. Accessing Minislots

In the distributed MAC, every voice node needs to exchange information with neighbor voice nodes by
broadcasting control packets in the minislots in the control period of each superframe. When the network
operations are start, the number of minislots (voice capacity Nvm) in the control period should be determined
in a distributed way under the constraint that voice packet loss rate is bounded by PL and the summation of
Tctrl and Tcfp does not exceed ‘ _ TSF in each superframe. After Nvm is determined, each voice node
randomly chooses one minislot in the CTP of a superframe1 and broadcasts a control packet in its selected
minislot [32]. Each node broadcast its control packet in the same occupied minislot of each subsequent
superframe2 until it is powered off or departs from the network. A control packet, shown in Fig. 2, includes five
fields: a header, a set of IDs of the node’s neighbors including the node itself, the node’s occupied minislot
sequence number (MSN, chosen from 1 to Nvm), buffer indication bit (BIB), the node’s scheduled TDMA slot
sequence number (SSN, a number within 0 to Ns) in the previous superframe.

Fig. 2 Format of control packet broadcast in each minislot

Accessing a minislot from a tag voice node is considered successful if the control packets received from other
nodes in subsequent minislots contain the tagged node’s ID [34]. Otherwise, an access collision happens due to
simultaneous control packet transmissions in the similar minislot by more than one node. All nodes are involved
in the collision wait until the next superframe to re-access one of the vacant minislots.

The minislot accessing process is completed when all existing nodes are successfully acquire their respective
minislots. When a new node is powered on or entering the network coverage area, it first synchronizes in time
with the start of a superframe, determines the number of minislots (based on '), Nvm, and listens to all control
packets in the CTP. Then, it randomly selects an unoccupied minislot and broadcasts a control of subsequent
superframes to broadcast its control packet.

Volume 5, Issue 10, October/2018 Page No:1316

JASC: Journal of Applied Science and Computations ISSN NO: 1076-5131

B. M- Lion Whale Optimization algorithm

Packet in the minislot in the next super frame. If all Nvm minislots are occupied, which means the entire
network reaches its voice capacity, the node defers its channel access and keeps sensing the CTPs of subsequent
superframes until some existing minislots are released due to node departures. After the minislot accessing is
successful, the node keep using the same minislot.This section elaborates the proposed multi-objective routing
approach in MANET using the hybrid optimization model, M-LionWhale. This multi-objective programming
model integrates LA [25] into WOA [19] for the optimal selection of the route for secured transmission. The
algorithm considers several QoS parameters, namely energy, distance with a security parameter, trust, as the
multi-objective or goal model. Once the algorithm measures the QoS and security parameters, it discovers k
possible paths from the source node to the destination node. The objective of the M-Lion Whale algorithm is to
select an optimum path from the discovered path, satisfying the multiple objectives considered for a secured
routing. The solution encoding, multi-objective fitness, along with the detailed explanation of the proposed
algorithm are given in the following sections. Fig. 3 depicts the block diagram of the proposed M-LionWhale
algorithm in secured MANET routing.

Fig. 3 Block diagram of the proposed M-LionWhale algorithm for secure routing

The above picture is taken from reference of “ M-LionWhale: Multi-Objective Optimization model for secure
routing in mobile ad-hoc network” An important element in presenting the algorithm is solution encoding. The
proposed approach of finding the optimal path from the source S to the destination D is considered as a binary
search problem. Let D = p1, p2, …, pk represents the destination having k paths formed by the intermediate
nodes between the source and the destination nodes. The number of paths in the network depends on the
number of nodes N, such that the range is 0 ≤ pk ≤ N. Hence, the number of solutions that exists in the
network between the source and the destination can be represented as SE = 1, 2, …, k , where k is the total
number of paths. From the k number of solutions that exist between the source-destination pair, the algorithm
finds a path such that the path is optimum with maximum trust and other preferred objectives using a
multiobjective fitness function, as explained in the section below.

Multi-objective fitness formulation

One of the essential factors that decide the worth of the solution is the fitness. The fitness function of the
proposed M-Lion Whale algorithm is formulated using goal model having five objectives, such as trust, energy,
delay, distance, and link lifetime. The solution having maximum trust, increased energy, reduced delay, minimum
distance, and higher lifetime, make the fitness maximum and thereby, forms the best solution. The fitness of the
proposed routing algorithm is defined as

Volume 5, Issue 10, October/2018 Page No:1317

JASC: Journal of Applied Science and Computations ISSN NO: 1076-5131

where Tj is the trust provided in a node j, Ej is the energy that exists in the node, dj, j + n is the distance between
two nodes j and j + n, Q is the threshold that adjusts the distance to be in minimum. NP is the number of nodes
in a particular path, N is the total number of nodes available, TLLT is the lifetime of a link between two nodes in
NP and A is a variable that normalizes TLLT. The objectives trust, energy, link lifetime, and distance are estimated
using the equations formulated in the corresponding models, as in Sections 3.1–3.4, respectively. The nodes
having maximum trust are selected as they ensure security during the transmission. The value of trust Tj ranges
between 0 and 1, where 0 is the best value and 1 is the worst. Similarly, the nodes with maximum energy and
link lifetime influence the fitness and are preferred for the formation of routes. Ej provides a value within [0,1],
where the nodes having a value 1 are selected to form the route. Likewise, the link lifetime TLLT and the distance
dj, j + n have a value between 0 and 1, where 0 forms the best value. The delay in the transmission is measured
as the ratio of the number of nodes present in a particular path to the total available nodes in the network, as
given in (17). This measure too has a value that ranges from 0 to 1, with 0 as the best and 1 as the worst value.
The nodes that satisfy all these objectives are selected to form the routes that exist from the source to the
destination. Based on the fitness formulated, the proposed M-Lion Whale approach, deliberated in the following
section, determines the best solution.

M-LionWhale-based goal programming approach for secure routing

The proposed hybrid optimization approach, named M-Lion Whale, for routing with security awareness is
illustrated in this section. M-Lion Whale algorithm is designed by the integration of LA into WOA in the update
rule. The proposed technique utilizes a goal model that considers four QOS metrics, energy, delay, distance, link
lifetime, and the parameter trust, as the objectives for the optimal selection of solution offering security. LA is
an optimization algorithm developed based on lion's social behavior, like territorial defense, takeover,
exploitation of laggard lions, and pride. WOA is a meta-heuristic optimization algorithm that is developed by
mimicking the humpback whales’ behavior of bubble-net hunting. Adopting the update rule of LA in WOA, M-
Lion Whale approach can develop an optimum solution with the security provided using QOS-security based
objectives. Algorithm presents the proposed M-Lion Whale multi-objective algorithm for secure routing.

IV.RESULT AND DISCUSSION

Network Simulator 2(NS2) was developed by UC Berkeley and it is written in OTcl and C++. NS2 is a discrete
event driven network simulator and it is primarily useful for simulating local and wide area networks. The
simulation results are represented as graphs for various performance analysis such as Source Frequency, Channel
Frequency, Destination Frequency and Packet Delivery Ratio. Simulation parameter

Volume 5, Issue 10, October/2018 Page No:1318

JASC: Journal of Applied Science and Computations ISSN NO: 1076-5131

In Fig.4 , graph represents the number of crests of a wave that move at a given point in a given unit of time.
Common unit of frequency is the Hertz (Hz) corresponding to one crest for each second. The frequency of a
wave can be calculated by dividing the speed of the wave by at sender side of wavelength

Fig. 4 Source Frequency

The following graph represents the destination frequency and it can be calculated by dividing the speed of the
wave by wavelength at the receiver side.

Fig. 5 Destination Frequency

Volume 5, Issue 10, October/2018 Page No:1319

JASC: Journal of Applied Science and Computations ISSN NO: 1076-5131

The Channel frequency of the wireless communication is represented in the Fig 6. X-axis represents the channel
frequency and Y – axis represents the time in m-sec. It focuses particularly on radio frequency, signal strength
and also field strength.

Fig. 6 Channel Frequency

In the Fig 7 graph is represent throughput is calculated using awk script which processes the trace file and
produces the result. Packet delivery ratio is the ratio of packets that are successfully delivered to a destination
compared to the number of packets that have been sent by sender. In order to calculate packet delivery ratio we
need total number of packets sent and number of received packets.

Fig. 7 Node Packet-Delivery Ratio

V.CONCLUSION

In this paper, we propose a distributed and traffic-adaptive hybrid MAC scheme for IoT-enabled MANET and
with secure routing in MANET using M-LionWhale algorithm. This hybrid optimization algorithm integrates
LA into WOA for the optimal selection of a path for the transmission. Trust is the security parameter
considered in MLionWhale algorithm that provides security based on the trust degrees of nodes. A hybrid MAC
superframe structure is devised, in which voice nodes are allocated a time slot in a distributed way by adapting to
their instantaneous transmission buffer states and data nodes contend to access the channel in a contention
period of each superframe according to the T-CSMA/CA. It will reduce the delay in data transfer, will increase
the data rate of packet transfer and routing delay will be reduced and data recovery will be possible.

Reference

[1] Ram Mohan Chintalapalli, Venugopal Reddy Ananthula “ M-LionWhale: Multi-Objective Optimisation model for secure routing
in mobile ad-hoc network” IEEE-2018
[2] Qiang Ye and Weihua Zhuang “Distributed and Adaptive Medium Access Control for Internet-of-Things Enabled Mobile
Networks” IEEE-2016

Volume 5, Issue 10, October/2018 Page No:1320

JASC: Journal of Applied Science and Computations ISSN NO: 1076-5131

[3] Qiang Ye and Weihua Zhuang “Token Based Adaptive MAC for Two-Hop Internet-of-Things Enabled MANET” IEEE -
2016
[4] H. Nishiyama, T. Ngo, S. Oiyama, and N. Kato, “Relay by smart device: Innovative communications for efficient information
sharing among vehicles and pedestrians,” IEEE Veh. Technol. Mag., vol. 10, no. 4, pp. 54–62, 2015.
[5] J. Liu and N. Kato, “A markovian analysis for explicit probabilistic stopping-based information propagation in postdisaster ad hoc
mobile networks,” IEEE Trans. Wireless Commun., vol. 15, no. 1, pp. 81–90, 2016.
[6] J.-R. Cha, K.-C. Go, J.-H. Kim, and W.-C. Park,“TDMA-based multihop resource reservation protocol for real-time
applications in tactical mobile adhoc network,” in Proc. IEEE MILCOM’10, 2010, pp. 1936–1941.
[7] Marimuthu, M., Krishnamurthi, I.: ‘Enhanced OLSR for defense against DOS attack in ad hoc networks’, IEEE Journal of
Communications and Networks, 2013, 15, (1), pp. 31–37
[8] Muthusenthil, B., Murugavalli, S.: ‘The impact of location-based attacks on geographical routing protocols', J. Theor. Appl. Inf.
Technol., 2014, 60, (2),pp. 189–199
[9] Hussain, Z., Balakrishna, R.: ‘A survey on manets – types, characteristics, applications, and protocols used'. National Conf. on
Frontiers & Advances in Information Science and Technology, 2014
[10] Mohanapriya, M., Krishnamurthi, I.: ‘Modified DSR protocol for detection and removal of selective black hole attack in
MANET’, Comput. Electr. Eng.,2014, 40, (2), pp. 530–538
[11] Chang, J.M., Tsou, P.C., Woungang, I., et al.: ‘Defending against collaborative attacks by malicious nodes in
MANETs: a cooperative bait detection approach’, IEEE Syst. J., 2015, 9, (1), pp. 65–75
[12] Shabut, A.M., Dahal, K.P., Bista, S.K., et al.: “Recommendation based trustmodel with an effective defence scheme for
MANETs’, IEEE Trans. Mob.Comput., 2015, 14, (10), pp. 2101–2115
[13] Mirjalili, S., Lewis, A.: ‘The Whale optimization algorithm’, Adv. Eng. Softw.,2016, 95, pp. 51–67
[14] Moustafa, H., Labiod, H.: ‘Multicast routing in mobile ad hoc networks’,Telecommun. Syst., 2004, 25, (1-2), pp. 65–
88
[15] Chander, S., Vijaya, P., Dhyani, P.: ‘Fractional lion algorithm – an optimization algorithm for data clustering’, J.
Comput. Sci., 2016, 12, (7), pp. 323–340
[16] Ali, H., Shahzad, W., Khan, F.A.: ‘Energy-efficient clustering in mobile adhoc networks using multi-objective particle
swarm optimization’, Appl. Soft Comput., 2012, 12, (7), pp. 1913–1928
[17] Rajan, C., Shanthi, N.: ‘Genetic based optimization for multicast routing algorithm for MANET’, Indian Acad. Sci.,
2015, 40, (8), pp. 2341–2352
[18] Dand, P., Radhakrishnan, S.: ‘An ant colony-based multi objective quality of service routing for mobile ad hoc networks’,
EURASIP J. Wirel. Commun.Netw., 2011, 153, pp. 1–12
[19] Rajakumar, B.R.: ‘Lion algorithm for standard and large scale bilinear system identification: a global optimization
based on Lion's social behavior’. 2014 IEEE Congress on Evolutionary Computation (CEC), Beijing, China, 2014,pp. 2116–
2123

Volume 5, Issue 10, October/2018 Page No:1321