2018 5th IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and Computer Engineering

(UPCON)

QOS Improvement in MANET Routing by Route

Optimization through Convergence of Mobile Agent
Vandna Rani Verma D.P. Sharma C.S. Lamba
Dept of M.C.A, MAISM, Jaipur Research Centre, MSRDC, jaipur Dept of Computer Engineering
Rajasthan Technical University, R.T.U, Kota , India Manipal University,Jaipur, India
Kota, India AMIT, AMU, MOEFDRE, UNDP
vandna.mtech@gmail.com

Abstract—A, Mobile ad-hoc network (MANET) is collection of congestion in network. Node mobility and energy are the two
mobile devices that desire to communicate in the lack of any fixed main factors of link breakage.
infrastructure so, in this environment every mobile nodes works
as a router in itself. In this type of wireless network, any node can
leave or join network at any movement. Besides it this type of
networks have some other challenging characteristics as
regularly changing topology, high mobility, narrow bandwidth
.Conventional routing protocols focused on finding shortest path
rather than Quality of Service path. Link breakage is the main
reason of high mobility as well as lack of energy of wireless nodes.
In the lack of resources, traffic density leads to congestion. So, the
performance of routing protocol degrades in this type of Fig- 1. An infrastructure- less wireless network
networks. But it is not so easy task to take multiple QoS
parameters for finding optimal route. Mobile software agent
technology with unique features like mobility, autonomous, A. Routing Mechanism in MANET
intelligent and adaptability can solve this problem. This research The primary purpose of a routing mechanism in MANET is to
paper focuses on designing routing model using mobile agent find the path with the constraint of uniqueness of MANET
paradigm. The proposed model attempts to find more optimal
networks. Routing protocols fall into two main categories:
path by considering multiple QoS parameters like stability, delay,
congestion and energy of the node. Simulation is done using firstly, Reactive or on demand routing protocols, (e.g., Ad Hoc
Matlab simulator & the simulation output shows that our On-Demand Distance Vector (AODV), Dynamic Source
proposed model is more reliable than traditional hop by hop Routing (DSR) and secondary, Proactive or Table driven
AODV for finding the route from source node to destination Routing protocols (e.g., OLSR (Optimized Link State
node. Routing) [2]. Routing overhead decreases in reactive routing
protocols in comparison of proactive routing protocols because
Keywords— Quality of Service (QoS), Mobile ad-hoc Network in reactive routing techniques, mobile nodes search routing
(MANET), Mobile agent (MA), Ad-hoc On demand Distance paths only when source sends a request to find path for
vector(AODV).
unknown destination. But in proactive routing protocols,
I. INTRODUCTION. topology information of wireless nodes exchange regularly so
every time route information is available and a large capacity
Mobile Ad hoc network is an Infrastructure-less wireless
of network used. In this paper we focused on only AODV,
network and it does not require any fixed infrastructure, such
reactive routing protocol [2]. It is a uni-cast, self-initiated and
as a base station, for their operation [1]. In general, routes
dynamic routing that permits the mobile nodes to react to link
between nodes in an ad hoc network may include multiple
breakage. In this loop–free routing, route discovery
hops, and hence it is appropriate to call such networks as
mechanism starts when a route is needed from source node S
‘‘multi-hop wireless ad hoc networks’’. Each node will be able
to a unknown destination node D to send data, a route request
to communicate directly with any other node that resides
(RREQ) message is broadcasted to its all neighbors.
within its transmission range [1]. Fig-1 shows an
Destination or any intermediate node can reply who knows the
infrastructure-less wireless network communication in which
path to this unknown destination. AODV maintains three
any node can communicate within its transmission range In
control messages RREQ (Route Request), RREP (Route
MANET, to maintain network continuous, efficient and stable,
Reply, RERR (Route Error) for finding route and maintaining
it is necessary to take in to account the routing nodes load and
the MANET network.

978-1-5386-5002-8/18/$31.00 ©2018 IEEE

 B. Role of Mobile Agent in MANET Routing which transfers to server side and build local communications
To avoid link failure and for reducing communications with it [4, 6]. So, application of Mobile agent model is widely
overhead; stable routing is needed. However many critical used in distributed computing.
parameters like stability, flow density, energy and reliability D. Benefits of using Mobile AgentMobile Technology
have not been considered together or are taken with insufficient
details in existing researches. Taking single routing objective
such as increasing packet delivery ratio or increasing the life Reducing network load & network latency: Mobile agents can
time of the network can severely compromise network. reduce network load and latency by running remotely. Mobile
Therefore, efficient routing requires selection of route that meet Agent is mainly used for network management and distributed
multi objective in Mobile Ad-hoc networks. But taking into computations.
consideration multiple parameters in routing is an NP- Adaptability: Adaptability is very important feature used in
Complete problem [6, 13]. The complexity of routing protocol a environment where topology of network continuously
due to its dynamic nature can be reduced with convergence of changing or other environment features can change
mobile software agent because mobile agent can move to any continuously so, Mobile agent has intelligence to response
node and update the routing table and used to transfer the according to this change.
processing task to the data source to reduce the network load
[5]. Due to its unique feature of adaptability, it can maintain  Reduced Communication Cost: Mobile agent technology
dynamic topology of the network and can deduce link failure. can reduce communication cost by allowing remote
The complexity of routing protocol due to its dynamic nature communications to activate like local communications.
can be reduced with convergence of mobile software agent  Asynchronous execution: Mobile agent can migrate from
because mobile agent can move to any node and update the one node to another and there is no need to communication
routing table and used to transfer the processing task to the data with its source side node after migrating to destination side
source to reduce the network load [6] and due to its unique node.
feature of adaptability, it can maintain dynamic topology of the  Autonomous: Mobile Agent has its own whole control
network and can deduce link failure. of thread and enough information to make a decision
where and when to go [4,5] .
 Dynamic-deployment of software: Many applications of
C. Mobile Agent Pradigm& its Application Areas network management and distributed computing need
Mobile Agent can widely used in Network management, dynamic deployment of software.[6]
dynamic computing, E-commence and many other
applications. It is very beneficial for network management II. LITERATURE REVIEW
because it can reduce the network bandwidth [4, 5].
Literature work done in area of stability, load balancing and
energy QoS metrices .Using mobile agent technique. In [8]
,authors improve the performance of AODV against
congestion using mobile agent with load balancing scheme
using Medium Access Control drops and interface queue size.
In [9] ,congestion status is updated using mobile agent and
congestion status is calculated by buffer queue length ,In [10] ,
authors proposed a innovative routing Mobile Agent-AODV,
and proportion of number of route exceeding a node to the
Fig -2(a): Client/Server model
number of available route in the network .In, [11], a multi
agent routing is proposed to mitigate congestion.In [12] , a
reliable path is constructed by mobile agent based proposed
routing model taking three QoS metrices current load, packet
forwarding ratio of nodes and bandwidth etc. in [13],
Intelligent Ant agent based routing decision is taken by
considering stability based on link expiration time combined
with energy, delay and in [19], Flow control & for transmitting
Fig -2 (b): Mobile Agent Based model the packet , maximum life time and flow density is considered
.In [14], fuzzy crispy inputs are taken for two most important
issue for reliable routing stability and energy in this proposed
Fig-2a & 2b show the basic dissimilarity between client-server routing multiple metrices for stability of link are taken in to
model and mobile agent based model. Communication cost consideration with residual energy and monitoring agent is
and overhead increase in client server model due to a number used for selection of routing path based on multipath routing.
of communications between client and server in shape of In [15], AODV routing protocol improved using multiple QoS
requests messages and replies messages. So, reducing this issues like frequent link breaks, delay, and un -balanced load
network load, the idea of mobile agent is utilized. In mobile of nodes with varying network size and definitely this routing
agent design, the client communicates with only mobile agent proposal provides Quality of service with respect to delay
metrices but it lacks in adaption and intelligent decision in for modified packet of RREQ. QoS metrics link stability is
finding route. In [7], authors proposed a mobile agent based calculated using RSS (Received Signal Strength), congestion
power and congestion aware routing in which agent starts to is measured by the ratio of number of packets in the queue to
send the data to the route having min cost in terms of energy buffer size & Remaining energy is achieved by detecting
cost and congestion. Nodes can be divided in listening and non consumed energy with respect to total energy. Queuing delay
listening mode. In [21], authors proposed adaptive mobility is considered as the waiting time in the queue for a packet
aware AODV based routing protocol and this protocol takes before it is sent out onto a link. Let n1, n2, n3…………ni, are
velocity, hop-count and length of the queue for selection of the neighbours of source node Si. Si sends data packet based
path. In [20], authors modified traditional routing AODV with on PSI (Path Selection Index).
load-distribution metric aggregate interface queue length for N
PSI max= 1/N ∑ (LSi ×REi× (1-CLi) × (1-Qdelayi))……………. (1)
selecting less loaded path for sending data. A detailed survey
i=1
on MANET routing implementation were covered using Where, N represents total number of mobile nodes in selected path.
neighborhood mechanism, hash tables, various traffic models
and bloom filter [23]. In [24] authors presented a mobile sink
as checkpoints to detect fault in sensor network with fuzzy Type J R G U Reserved Hop-count
rule based better recovery probability in comparison of
random checkpoint. A fuzzy rule based check point detection
scheme presented for mobile grid [25]. a cluster based RREQ ID ( Route Request ID)
Source IP
approach is used to find malicious , critical as well as popular Source sequence number
nodes [26]. In [27] authors proposed a fault detection Destination IP
technique based on fuzzy logic based with three QoS Destination sequence number
parameters 1.e. link quality, hop-count & received signal PSI
strength for wireless sensor network .

Fig- 4. Route Request format of proposed routing model.

III. PROPOSED SYSTEM
RREQ packet format of proposed model is given in Fig -4.
The main objective of this system is to discover the best
RREQ contains Route request ID (RREQ ID), Source IP,
possible path for improvement in Quality of Service of
Source Sequence address, destination IP and destination
MANET routing with convergence of Mobile Software Agent
sequence number and PSI (Path Index Value). Simulation
(MSA) by taking multiple Quality of Service (QoS) metrices
Based Study & Results In this simulation work, inputs are
like link stability, congestion, delay & remaining energy . This
defined in table-1. Simulation is done on proposed routing
work is based on modified version of AODV routing protocol
model and hop by hop AODV (Ad-hoc on-demand Routing)
and using mobile Agent concept [29, 30].
using matlab2017a. In simulation input parameters like
number of mobile nodes, number of source node as well as
Proposed Routing Model
destination node, X and Y co-ordinates of mobile nodes,
mobile node transmission range & node trust values are taken
in to consideration. Node trust values for node stability,
remaining energy, delay and congestion are to be considered
within range from 0-1. Number of nodes and well as node
positions can be changed.

TABLE I. INPUT PARAMETERS

Input Parameter `Value

Total number of nodes used 10
Fig - 3. Mobile Agent Based stable energy proficient and load balanced Index of source node 1
routing model for MANET
Index of source node 10
According to Fig-3, the source node broadcast RREQ packet Node transmission range 5
to its one hop neighbours and this RREQ packet select the
Node X-coordinates 1, 2, 3, 4, 8, 6, 7, 9, 10 ,10
shortest path towards destination. To travel towards
destination node D, routing information is updated using Node Y-coordinates 6, 2, 5, 8, 5, 1, 10 ,2 ,8 ,5
RREQ packet. Now source node 1 selects the path with max Nodes stability value 1 ,1.0 ,1.0 ,1.0 ,1 ,1, 1.0 ,1, 1, 1
PSI cost. RREQ packet moves towards Path P1, P2……hop
by hop manner is broadcasted to its all one-hop neighbours Nodes delay value 1, 1.0, 1.0, 1.0, 1, 1 ,1.0 ,1, 1, 1
then to forward RREQ to its intermediate nodes it uses
 Nodes congestion value 1 ,1.0, 1.0, 1.0, 1, 1 ,1.0, 1 ,1, 1 Path hopbyhop_trusted_path antPSI_trusted_path_distance
Distance distance-26.5253 - 9.1623
Nodes Remaining Energy .3, 1.0, 1.0, 1.0 ,1 ,1 ,1.0, 1 ,1, 1
In the above Table-2, a comparative analysis on the basis
of simulation is done between hop by hop AODV model and
TABLE II. TABLE STYLESCOMPARISION OF HOP BY HOP AODV MODEL
WITH PROPOSED MODE proposed model. Table-2 shows the comparative analysis on
the basis of different QoS parameters like path cost, path hop
Parameters (Hop by Hop) routing Proposed model and path distance.Table-2 shows that proposed model
AODV
Selected hopbyhop_trusted_path – antPSI_trusted_path
performed is better than hop by hop AODV model. Proposed
Routing 1-2-3-4-5-6-8-10 1 3 5 10 model selected fewer hops and less path distance from source
Path to destination node. In,Fig-5 shows calculated path distance,
Path Cost hopbyhop_cost - antPSI_cost - path cost, and path hop of hop by hop AODV routing and Fig-
0.8750 0.6667 6, shows the values of path selected by
Path Hop hopbyhop_trusted_path_ho antPSI_trusted_path_hops-
ps -7 3

Fig- 5. Snapshot of output of QoS Parameters of Hop by Hop AODV routing model i.e. Selected Routing Path, Path Cost, Path Distance, Hop.
 Fig - 6. Snapshot of output of QoS Parameters of proposed routing model like Selected Routing Path, Path Cost, Path Distance, Path.

Fig-7: Snapshot of comparison of shortest path selected between hop by hop AODV routing and proposed routing model.

Proposed routing model i.e. selected path, path cost, path we will add more results like computation time of proposed
distance, path hops etc source to destination. In Fig-7, snapshot routing model and also try to apply this proposed solution on
shows two different routing paths. One path is indicated using other reactive routing protocol like Dynamic Source Routing
red colour line selected using hop by hop AODV routing and (DSR) .
other path is indicated by using green line selected by our References:
proposed routing model. From Fig-7, it is clear that path
selected from proposed model indicated by green line is much [1] Ram Ramanathanand , Jason Redi “A brief overview ad-hoc networks”
optimized than hop by hop traditional routing indicated by red IEEE Communications Magazine 50th Anniversary issue .
line in terms of path distance, path hop to reach to destination. [2] J.Raju ,j.j.Garcia ,” a comparation of on-demand and table driven routing
So, we can increase the battery life time of the node by for ad-hoc wireless networks,”, Ieee international conference.
[3] C. E. Perkins, "Ad -Hoc On Demand Distance Vector (AODV) Routing,"
choosing shorter route. Also shows the path selected by hop by IETF Internet Draft, RFC 3561.
hop AODV and proposed routing algorithm. [4] M. L. Griss , G. Pour, “Development with Agent Components., IEEE
Computer magazine, vol.34, no.5, pp.37-43, 2001.
[5] N. R. Jennings, “An Agent-Based Approach for Building Complex
software systems “., Communications of ACM, vol.44, pp.35-41,. 2001.
IV. CONCLUSION [6] Minar N., Kramer K. H.,& Maes P., Cooperative Mobile Agents for
The main objective of this proposed model is to discover the Dynamic Network Routing, in Software Agents for Future.
best path on the basis of combined multiple QoS parameters [7] Sharma V, Bhadauria SS,” Congestion and power control techniques based
on Mobile agent and varying rate in MANET, International journal of
stability with consideration of energy, load balancing and advancements in computer science and information technology[2250-2440],
delay. We simulated the proposed model using Matlab 2017 & vol. 01, no. 1, sep ,2011, pg 24-34.
experimental results concluded that proposed improved model [8]Shekhar H M P, K S Ramanatha, "Agent Based Congestion Aware Routing
is much efficient than hop by hop AODV routing on the basis in Mobile Ad Hoc Networks”, Technical Report, Wireless networking Lab,
Department of Computer Science & Engineering, M S Ramaiah Institute of
of three parameters i.e. path cost, hops, and path distance etc Technology, India, 2005.
for finding path. So, output of experimental results shows that [9] Hong Li, Chu Dan, Wang Min, Li Shurong,”Mobile Agent Based
proposed model is much more optimized for these three Congestion Control AODV Routing Protocol” Wireless Communications,
parameters compared than hop by hop AODV routing can be Networking and Mobile Computing, 2008. IEEE,WiCOM '08. 4th
International Conference, Dalian,12-14 Oct. 2008.
seen for finding path from source to destination. So, with the [10]CungTrongCuong et al. “MAR-AODV: Innovative Routing Algorithm in
selection of shortest path we can save much more battery of MANET Based on Mobile Agent “WAINA '13 Proceedings,27th International
the mobile nodes and can avoid network partition because Conference on Advanced Information Networking and Applications
mobile nodes has limited battery power & many nodes is not Workshops Pages 62-66 IEEE Computer Society , 2013 ISBN: 978-0-7695-
4952-1 .
able to forward path in the lacking of battery power. In future,
[11] Kazuya Nishimura and Kazuko Takahashi, “A Multi-Agent Routing
Protocol with Congestion Control for MANET”, Proceedings 21st European
Conference on Modeling and Simulation, 2007.
[12] G. Varaprasad “Stable routing algorithm for mobile ad hoc networks
using mobile agent “,”, International journal of communication system “,2012.
[13]M. Belkadi et al.” Intelligent Routing and Flow Control in MANETs ”
Journal of Computing and Information Technology - CIT,vol. 3, pg,233–243,
[14] Palaniappan & Chellan” Energy-efficient stable routing using QoS
monitoring agents in MANET” Journal on Wireless Communications and
Networking springer, 2015, pg 1-11.
[15] Mandeep Kaur Gulati &Krishan Kumar “Stable Energy efficient QoS
based Congestion and Delay aware Routing (SEQCDR) Protocol for
MANETs, IEEE ICCSP conference, 2015,pg1-7.
[16] PhateN,Saxena M & Rizvi M A “Minimizing Congestion and Improved
QoS of AODV using Clustering in Mobile Ad Hoc Network “, IEEE, Recent
Advances and Innovations in Engineering (ICRAIE), may, 2014,pg 9-
11.(ieee,2014).
[17] Sharma DP &Gautam Sumit, Distributed and Prioritised Scheduling to
Implement Carrier Aggregation in LTE Advanced Systems, International
Conference on Advanced Computing & Communication Technologies
(ACCT), 2014.
[18]Zhu Qiankun et .al “A Mobile Ad Hoc Networks Algorithm Improved
AODV Protocol” International Conference on Power Electronics and
Engineering Application(PEEA 2011) Elsevier Science Direct , 2011, pg, 229
– 234.
[19] Jung et. Al. “Load Balanced Congestion Adaptive Routing for Mobile Ad
Hoc Networks” International Journal of Distributed Sensor Networks
Volume 2014 (2014), Article ID 532043.
[20]Amita Rani and Mayank Dave “Performance evaluation of modified
AODV for load balancing,”Journal of computer science, vol-3, pp 863-
868,2007.
[21]BisengerAhmed, Rziza Mohamed, Ouadou Mohamed “Queue Length and
Mobility aware Routing Protocol for Mobile Ad hoc Network” vol4 No-3,
2012, pg ,207-2011.
[22] Idrees M et al. “Enhancements in AODV Routing using mobility aware
agents” IEE, sep 17-18, 2005,pg 98-102.
[23] Renisha P. Salim ; Rajesh R” survey: Optimal node routing strategies in
MANET” IEEE Conferences International Conference on Data Mining and
Advanced Computing (SAPIENCE),2016 Pages: 260 – 267.
[24]Pritee Parwekar, Sireesha Rodda,Parmeet kaur,"Mobile Sink as
Checkpoints for Fault Detection towards Fault Tolerance in Wireless Sensor
Networks”, Published in Journal of Global Information Management (JGIM),
volume 26 issue 3. doi:10.4018/JGIM.
[25]Kaur, P.; Parwekar, P, "Fuzzy rule based checkpointing arrangement for
fault tolerance in Mobile Grids". Published,Contemporary Computing (IC3),
2014 Seventh International Conference on DOI:10.1109/IC3.2014.6897188
Publication Year: 2014,Page(s):289- 293.
[26]Parwekar, P., & Arora, S. (2014). Security Issues and Its Counter
Measures in Mobile Ad Hoc Networks. In ICT and Critical Infrastructure:
Proceedings of the 48th Annual Convention of Computer Society of India-Vol
I (pp. 301-309). Springer, Cham
[27]Pritee Parwekar, Sireesha Rodda,”Fault Tolerance in Wireless Sensor
Networks: Finding Primary Path”, Published in Proceedings of the Second
International Conference on Computer and Communication Technologies,
593-604, Springer India publication,2016
[28] Matlab documentation “https://in.mathworks.com/help/matlab/ “.
[29]Vandna Rani Verma, D. P. Sharma, C.S Lamba "Stable energy proficient
and load balancing based QoS routing in mobile Ad-Hoc networks: Mobile
software based approach" Malaya Journal of Matematik, vol-1, issue-1, 2018.
[30] Vandna Rani Verma, D. P. Sharma, C.S Lamba “ load balancing and
stability based Quality ofService in adhoc networks: Software Agent based
approachInternational Conference on Research and Innovations in
collaboration with United Group & Asian Institute of Technology Thailand
vol1 2016 pg, 70-75