Likelihood based Node Fitness Evaluation Method for Data Authentication in MANET

Abstract

Mobile nodes in ad hoc network are easily compromised by attackers. Due to the presence of
attackers, the network may be overloaded which leads to least security. In this research work,
Likelihood based Node Fitness Evaluation Method (LNFEM) is introduced and developed based on
trust model. The trust model consists of three phases. In first phase, trust model is defined based on
trust generation and computation. Node recommendation is used to produce the trust vector. In second
phase, the clustered secure routing is adopted to provide seamless connectivity in the presence of
attackers. In third phase, direct evaluation system is used to inter relate the direct and indirect
observation. The proposed method is evaluated using network simulation tool (NS2.3). The performance
metrics are throughput, data authentication ratio, data confidentiality rate, control overhead and
propagation delay. From the simulation results, the proposed method achieves better performance
compared to the existing methods.
Keywords Likelihood function, trust computation, trust generation, cluster, throughput, data authentication
ratio, data confidentiality rate and control overhead.

Full text

International Journal of Advanced Science and Technology
Vol. 29, No. 3, (2020), pp. 5835 - 5842
5835
ISSN: 2005-4238 IJAST
Copyright ⓒ 2019 SERSC
Likelihood based Node Fitness Evaluation Method for Data
Authentication in MANET
S. Pramela Devi*, Dr.V. Eswaramoorthy2, Dr. K. Vinoth Kumar3 and T.
Jayasankar4
1*Assistant Professor, Department of CSE, MVJ College of Engineering, Channasandra, Bangalore.
2Associate Professor, Department of ECE, SSM Institute of Engineering and Technology,Dindigul.
3Associate Professor, Department of CSE, SSM Institute of Engineering and Technology, Dindigul.
4Assistanr Professor, Department of ECE, Bharathidhasan Institute of Technology, Trichirappalli.
Abstract
Mobile nodes in ad hoc network are easily compromised by attackers. Due to the presence of
attackers, the network may be overloaded which leads to least security. In this research work,
Likelihood based Node Fitness Evaluation Method (LNFEM) is introduced and developed based on
trust model. The trust model consists of three phases. In first phase, trust model is defined based on
trust generation and computation. Node recommendation is used to produce the trust vector. In second
phase, the clustered secure routing is adopted to provide seamless connectivity in the presence of
attackers. In third phase, direct evaluation system is used to inter relate the direct and indirect
observation. The proposed method is evaluated using network simulation tool (NS2.3). The performance
metrics are throughput, data authentication ratio, data confidentiality rate, control overhead and
propagation delay. From the simulation results, the proposed method achieves better performance
compared to the existing methods.
Keywords Likelihood function, trust computation, trust generation, cluster, throughput, data authentication
ratio, data confidentiality rate and control overhead.
1.Introduction
In this modern world, handheld devices like laptops, mobile phones and tablets are very
important and play a major role in everyone’s life. Mobile ad hoc network is a kind of network
where it support vehicle networks very well. It is helpful in disaster management, emergency
applications, earth quake and so on. MANET is a powerful platform which provides connectivity,
mobility and flexibility to all devices in the world.
Security is an important concern in ad hoc networks due to weak signal strength, limited
physical resources and less protection of mobile nodes. Trust can act as important role to provide
the authentication of ad hoc networks. Due to bandwidth constraints, it is not at all possible the
participation of all nodes in the network. Both cryptography and trust model can be combined to
provide the entire security in the network.
2. Previous Work
Ankita Gupta and Abhishek Dubey [1] proposed the trust based approach to prevent
black hole attack using dynamic source routing protocol. In this routing, before packet forwarding
process, the entire route stability was found to prevent attackers inside the network. The secure
routes were established from source to destination by deploying the trust enhanced approach. The

International Journal of Advanced Science and Technology
Vol. 29, No. 3, (2020), pp. 5835 - 5842
5836
ISSN: 2005-4238 IJAST
Copyright ⓒ 2019 SERSC
multiple routes were found to provide reliable and efficient data transmission. The shortest routes
are found using effective algorithms which will improve the network performance dramatically.
Shanthi et.al [2] presented the weighted path mechanism to improve the quality of service
to reduce delay. The path speed determines the priority of the path between source to sink nodes.
The quality of the paths was determined by the weighted path procedure. The speed of the path
and network routes determines the path speed after effective packet transmission. It leads to the
reduction of delay, overhead and packet loss. The high dependability reduced the maximum packet
transfer and validation administration.
Niloofar Movahedian Attar [3] identified the secure routes in the ad hoc environment. The
on demand vector routing was discovered from source to sink node to support the prevention of
malicious nodes. The equation was derived to estimate the trust vector in order to find the secure
routes. The performance of the protocol was analyzed and summarized effectively during omnet
simulation environment.
Rajesh Yamparala and Balamurugan Perumal [4] developed the effective routing method
using group key management concept to ensure secure data transmission. The different qualities of
attack were found to provide authenticated nodes. The main aim of this protocol was to identify the
packet dropping attack using hybrid cryptography. The session keys for the nodes were arranged to
build secure trust model to sort out the several attackers during packet transmission.
Rohi Tariq et.al [5] proposed the secure on demand vector routing protocol to prevent the
black hole and denial of service attack. The hop count of routes was increased due to the presence
of malicious node. All route request packets towards the intermediate nodes are identified towards
the positive data transmission. The source node sends the packets to neighbor node without
knowing the presence of black hole attacks. These attackers were reduced by adopting crypto
mechanism.
Zulfiqar Ali Zardari et.al [6] proposed the intrusion detection system with connecting
domain set to balance energy and nonexistence of blacklisted nodes. The localized approach was
used to find the location of nearby connecting domain sets of nodes. The complete character of all
nodes can be retried using the selected intrusion detection system node to broadcast status of
packet. Finally the malicious node and black listed nodes were identified and removed using the
proposed intrusion detection system.
Perumal Raghavan Kavitha and Rajeswari Mukesh [7] developed the enhanced
polynomial reduction algorithm to tackle the security issues in the clustered ad hoc environment.
The strong key authentication based on two set of primitive polynomial procedure to identify the
effective routes in a short span of time. The neighbor routers are used to produce the
acknowledgements and also to identify the impersonation to reduce overhead.
Adwan Yasin and Mahmoud Abu Zant [8] developed the timer based baited technique to resist the
black hole attacks based on timers and baiting messages. The fake id was creased in the baiting
phase based on timer which reaches the threshold. The best paths were found based on behavior of
the attackers. The black hole attackers were identified and removed from the network using the
baiting technique. The TTL (Time-To-live) value was to isolate the congestion from the fake
requests.
Saranya and Padmapriya [9] proposed the system to reduce and prevent the both black
hole & grayhole attack in MANET grayhole attack based secure routing algorithm. The impact of
the attackers was identified and reduced using on demand routing. The control packet algorithm

International Journal of Advanced Science and Technology
Vol. 29, No. 3, (2020), pp. 5835 - 5842
5837
ISSN: 2005-4238 IJAST
Copyright ⓒ 2019 SERSC
was proposed to investigate the behavior of black hole and gray hole attacks. The elliptic curve
cryptography technique was used to reduce the impact of attacks to provide authentication.
Gopinath and Vinoth Kumar [10] presented the novel technique to detect the wormhole
attack. The introduced technique was the improvement of Delphi technique to detect and isolate
the malicious node. The per hop delay was estimated based on congestion and link failure. The
threshold delay was calculated to reduce the packet loss and misbehaving nodes.
3. Likelihood based Node Fitness Evaluation Method (LNFEM)
In this section, trust model is developed to find the trustworthy path and effective packet
delivery in ad hoc network. The concept of clustered network environment is determined to initiate
the stable routing. The direct reliability and indirect reliability metric are estimated to good and bad
nodes. There are three major steps to identify the good and false forwarding nodes.
1. The first one is number of data packets successfully forwarded without dropping from
source to destination.
2. The second step is to identify the packet dropping between neighbor nodes.
3. The third one is to detect the falsely injected packets in routing without the knowledge
of source and destination node.
Likelihood based node fitness evaluation method is proposed here to ensure data
authentication. It consists of three major models. The first model is trust model which ensures end
to end reliable packet delivery and availability of nodes. The second model is trust generation and
the third is trust computation model. Both models are used to ensure data and node authentication.
3.1 Trust Model for Revocation of Reliability
In this first model, cluster is formed with group of cluster members and the Cluster Head
(CH) is chosen based on node recommendation and node reputation. The node recommendation is
used to identify the reliable intermediate nodes in the cluster region. CH initiates the data
transmission process and the trade-off between reliability withdrawals and reliability starting
phase. The CH manager is created based on the voting by CH. This manager assigns aging metric
to each node which may hide the reliability value. It can be preset based on requirement of the
user. In MANET, packets may be dropped due to the network congestion, network contentions and
selfish as well as misbehaving nodes. Bad nodes are the nodes which drops the packets slowly.
CH divide the members into a group based on threshold factor. Based on the resultant vector of
each node, the CH broadcasts the trust vector value to the remaining cluster members.
The following steps to ensure the trust model of data delivery.
Intermediate node discovery:
In this phase, the list of trusted intermediate nodes is maintained along with reliability
metrics. In route discovery phase, end to end route is discovered from source to sink node with
good nodes. In route maintenance phase, each node is monitored by CH manager and it can
withdraw the route discovery process if the link failures occur unconditionally.
3.2 Generation of Node Trust
In this phase, the trust is calculated by CH and cluster members. If the lifetime of trust
value is expired, the trust cancellation procedure can be adopted to begin the trust calculation. Both
CH and cluster members are used to gather the behaviors of nodes based on route evidences. All
CH managers gather and record the data about good and bad event in its store table. The
probability distribution model is adopted to combine the feedback and node reputation.

International Journal of Advanced Science and Technology
Vol. 29, No. 3, (2020), pp. 5835 - 5842
5838
ISSN: 2005-4238 IJAST
Copyright ⓒ 2019 SERSC
3.3 Node Trust Calculation
In intermediate node management phase, CH collects the information about all nodes and
trust values of cluster members. It broadcasts beacon messages and store all node IDs in the store
table. Based on node recommendation, CH computes the trust value. The following steps are used
to identify the high trust value.
Step 1: CH broadcast the trust value to all neighbor nodes.
Step 2: The store table aggregator sends the beacon signal if the fake Id persists in the table.
Step 3: The likelihood function of node is estimated with zero mean and it is presented as,
( )

=




+=




22
2
2
log
2
1
,, tt
t
t
lBPml
Step 4: The likelihood of good behaving nodes is increased and trust cancellation begins properly.
3.4 Direct Evaluation System
In this phase, each cluster member can overhear the packets from the fake node based on
trust metric. The bayesian inference is used to build the direct evaluation system. Figure 1 shows
the illustration of direct evaluation system.
Application layer
Networking layer
Trust approach
Trust revocation
Trust Generation
Figure 1. Direct evaluation system
4. Simulation Results
The proposed method is evaluated using network simulator tool NS 2,35. The simulation
area size is 1200 x 1200 sq.m. The packet size of simulation is 256 bytes. Table 1. illustrates the
simulation settings and parameters.
Table1. Simulation and Setting Parameters of LNFEM
No. of Nodes
100
Area Size
1200 X 1200 sq.m
Mac
802.11
Radio Range
100 meter
Simulation Time
100 sec
Traffic Source
Poisson
Packet Size
256 bytes
Mobility Model
Random Way
Protocol
DSR
The following metrics are used to analyze the performance of proposed method.
Throughput: It is the number of packets transmitted per time.

International Journal of Advanced Science and Technology
Vol. 29, No. 3, (2020), pp. 5835 - 5842
5839
ISSN: 2005-4238 IJAST
Copyright ⓒ 2019 SERSC
Propagation delay: The number of packets propagated from source to intermediate nodes.
Control Overhead: It is the excessive number of packets over a period of time.
Data Authentication ratio: It is the number of nodes authenticated to the number of nodes
deployed.
Figure 2 illustrates the performance of proposed method in terms of throughput. Compared to
the existing schemes, the proposed scheme LNFEM achieves least throughput due to more traffic.
Figure 2. Throughput Vs No. of Packets
Figure 3 shows the analysis of propagation delay while varying time in x axis. Compared to the
existing schemes, the proposed scheme achieves less delay due to the computation of trust.
Figure 3. Propagation delay Vs Time
Figure 4 provides the analysis of Data authentication ratio. In x axis, the number of data packets
is varied. Due to the presence of data trust computation and generation method, the proposed
method achieves high ratio compared to existing schemes.

International Journal of Advanced Science and Technology
Vol. 29, No. 3, (2020), pp. 5835 - 5842
5840
ISSN: 2005-4238 IJAST
Copyright ⓒ 2019 SERSC
Figure 4. Data Authentication Ratio Vs No. of Data packets
Figure 5 shows the results of control overhead while varying the mobility in x axis. From the
results, it is seen that LNFEM achieves less control overhead than existing schemes.
Figure 5. Control Overhead Vs Mobility
Figure 6 shows the results of data confidentiality rate to ensure the protection of data from the
attackers. From the results, the proposed method achieves high rate than existing schemes.

International Journal of Advanced Science and Technology
Vol. 29, No. 3, (2020), pp. 5835 - 5842
5841
ISSN: 2005-4238 IJAST
Copyright ⓒ 2019 SERSC
Figure 6. Data Confidentiality rate Vs Speed
5. Conclusion
In the research community, Mobile ad hoc networks are the inseparable part of networks
where nodes are randomly deployed in the network. Due to dynamic nature, the reliability is
getting degraded. Trust evaluation is a major concern in the network. In this research work,
Likelihood based Node Fitness Evaluation Method is introduced based on the computation of node
trust. The cluster is formed with group of cluster members. The trust is estimated and distributed to
all intermediate nodes where the nodes use the trust vectors to forward the packets. Simulation
results are performed using network simulator tool. From the results, the proposed method
achieves better performance in terms of data authentication ratio, data confidentiality rate, control
overhead, propagation delay and throughput. In future, it is planned to implement IoT based secure
routing in MANET for real time applications.
REFERENCES
[1] Ankita Gupta, Abhishek Dubey, “ Advanced Technique Using Trust Based Approach for
Prevention of Black Hole Attack in Mobile Ad Hoc Network’, International Journal for
Research in Engineering Application & Management (IJREAM), Vol-04, Issue-02, May
2018, pp.84-89.
[2] Gopinath, S., Vinoth Kumar, K. & Jaya Sankar, T. Secure location aware routing protocol
with authentication for data integrity. Cluster Comput 22, 13609–13618 (2019).
https://doi.org/10.1007/s10586-018-2020-7
[3] P M. Shanthi, A. Dalvin Vinoth Kumar, L. Arockiam, “ A Technique to Enhance Quality of
Service using weighted Path Mechanism”, International Journal of Scientific Research in
Computer Science, Engineering and Information Technology, Vol.2, No.4, 2017, pp.413-
417.
[4] Niloofar Movahedian Attar, “Dynamic Detection of Secure Routes in Ad hoc Networks”,
Emerging Science Journal, Vol.1, No.4, 2017, pp.233-238.
[5] Rajesh Yamparala, Balamurugan Perumal, "Secure Data Transmission with Effective
Routing Method Using Group Key Management Techniques-A Survey", International
Information and Engineering Technology Association, Vol.52, No.3, 2019, pp.253-256.

International Journal of Advanced Science and Technology
Vol. 29, No. 3, (2020), pp. 5835 - 5842
5842
ISSN: 2005-4238 IJAST
Copyright ⓒ 2019 SERSC
[6] Rohi Tariq, Sheeraz Ahmed Corresp., Raees Shah Sani , Zeeshan Najam , Shahryar
Shafique, “Securing ad hoc on-demand distance vector routing protocol against the black
hole DoS attack in MANETs”, 2019, Peer J prints, 1-40.
[7] Zulfiqar Ali Zardari , Jingsha He , Nafei Zhu, Khalid Hussain Mohammadani , Muhammad
Salman Pathan , Muhammad Iftikhar Hussain and Muhammad Qasim Memon, “A Dual
Attack Detection Technique to Identify Black and Gray Hole Attacks Using an Intrusion
Detection System and a Connected Dominating Set in MANETs”, Future Internet, 2019,
pp.1-17.
[8] Perumal Raghavan Kavitha, Rajeswari Mukesh, “Detection of Impersonation Attack in
MANET Using Polynomial Reduction Algorithm”, International Journal of Network
Security, Vol.20, No.2, PP.381-389, Mar. 2018.
[9] Adwan Yasin and Mahmoud Abu Zant, “Detecting and Isolating Black-Hole Attacks in
MANET Using Timer Based Baited Technique”, Wireless Communications and Mobile
Computing, 2018, pp.1-11.
[10] K.Vinoth Kumar, T.Jayasankar, V.Srinivasan, .Prabhakaran, “Fuzzy Logic based Efficient
Multipath Routing for Mobile Ad-hoc Networks” Applied Mathematics & Information
Sciences(AMIS), ISSN: 1935-0090 (Print): EISSN: 2325-0399
(Online),vol.11,no.2,March 2017
[11] Saranya. R and Padmapriya. R, “Contradiction Based Gray-Hole Attack Minimization for
Ad-Hoc Networks”, International Journal of Advanced Research in Computer and
Communication Engineering, Vol. 7, Issue 2, February 2018, pp.8-14.
[12] K.Vinoth Kumar, Dr.V.Eswaramoorthy, S.Nagakumararaj and J.Wilson, “Fuzzy Clustering
Enchanced Multipath Routing to Enhance the Network Lifetime in Wireless Sensor
Networks”, International Journal Of Scientific & Technology Research Volume 8, Issue 11,
November 2019.
[13] K.Vinoth Kumar, Dr.S.Bhavani, “Trust Based Multipath Authentication Protocol for Mobile
Ad Hoc Network” Journal of Computational and Theoretical Nanoscience (CTN), ISSN:
1546-1955 (Print): EISSN: 1546-1963 (Online), Volume 12, Number 12, December 2015,
pp. 5479-5485(7) DOI: https://doi.org/10.1166/jctn.2015.4552.