Upload
others
View
7
Download
0
Embed Size (px)
Citation preview
http://www.iaeme.com/IJMET/index.asp 337 [email protected]
International Journal of Mechanical Engineering and Technology (IJMET) Volume 8, Issue 9, September 2017, pp. 337–348, Article ID: IJMET_08_09_036
Available online at http://www.iaeme.com/IJMET/issues.asp?JType=IJMET&VType=8&IType=9
ISSN Print: 0976-6340 and ISSN Online: 0976-6359
© IAEME Publication Scopus Indexed
MACHINE LEARNING TECHNIQUES FOR
WORMHOLE ATTACK DETECTION
TECHNIQUES IN WIRELESS SENSOR
NETWORKS
Er. Harpal
Research scholar, Shri Venkateshwara, University, Gajraula, India
Dr. Gaurav Tejpal
Professor, Shri Venkateshwara, University, Gajraula, India
Dr. Sonal Sharma
Assistant Professor, Uttaranchal University, Dehradun, India
ABSTRACT:
The wormhole attack in Wireless sensor networks (WSNs) decreases the network
performance by dropping the No. of Packets. Many techniques have been proposed to
so far reduce the impact of the wormhole attack by detecting and preventing it. But,
related work indicates that no technique is perfect for every kind of circumstances of
WSNs. Among the existing techniques, Watchdog technique has better performance in
preventing the wormhole attack. It utilizes the local knowledge of the next hop node
and eavesdrops it. If it gets that spending time of the Packet is more than the given
threshold, then it characterizes that node as wormhole attacker. However, this method
has several shortcomings that it does not track the link transmission errors, which
may be because of congestion in WSNs and also it not offers high mobility for
maximum No. of nodes, which eventually decreases the WSNs performance. In order
to handle this issue, a new multipoint relay based Watchdog monitoring and
prevention technique is proposed in this paper. Proposed technique utilizes the
dynamic threshold value to detect the wormhole attacker node, and then clustering
and the Watchdog based optimistic path is selected for communicating the Packets.
Thus, it reduces the overall Packet dropping, which improves the performance of the
WSNs.
Index Terms: black hole, wireless sensor networks, watchdog, multipoint relays.
Cite this Article: Er. Harpal, Dr. Gaurav Tejpal and Dr. Sonal Sharma, Machine
Learning Techniques for Wormhole Attack Detection Techniques in Wireless Sensor
Networks, International Journal of Mechanical Engineering and Technology 8(9),
2017, pp. 337–348.
http://www.iaeme.com/IJMET/issues.asp?JType=IJMET&VType=8&IType=9
Er. Harpal, Dr. Gaurav Tejpal and Dr. Sonal Sharma
http://www.iaeme.com/IJMET/index.asp 338 [email protected]
1. INTRODUCTION
Wireless sensor networks consist of individual sensor nodes, also called motes, deployed in a
given area that cooperatively collect and carry data to a main entity in order to monitor
physical or environmental conditions [1]. The main entity, also denoted as base station or
sink, can be connected to an infrastructure or to the Internet through a gateway, which allows
remote users to access the collected data [2]. The main advantage of wireless sensor networks
lies in the ability to deploy a lot of tiny autonomous motes without any pre-established
infrastructure [3]. After the deployment, motes gather information from the physical world,
and according to a defined communication protocol, they cooperate to deliver data towards
the sink through single-hop or multi-hop communications [4].
Figure 1 Typical wireless sensor network environment
The necessity for more effective security mechanisms for WSNs is increasing due to its
dynamic nature and continues growth in various fields. WSNs are organized in the
unfavorable environments. Different nodes in the WSNs have an unreliable communication
medium which makes it tough to deploy security mechanism [5]. Therefore, security of
different nodes in WSNs is a great challenge against various attacks. A variety of attacks are
possible in WSNs s including jamming, collision, wormhole, flooding, wormhole, sinkhole,
selective Packet drop, Sybil, cloning, denial-of-service, tampering etc. The Wormhole attack
is the most hazardous attack on WSNs [6].
Wormhole attack is the most hazardous attack in the network layer of WSNs. In this type
attack, malicious node acts as a wormhole in the universe [7]. A malicious node drops all of
the data Packets received from source node without transferring to the target node. In
wormhole attack, malicious node introduces itself as a node having the smallest route to the
destination node. Wormhole attack in WSNs is performed by an internal malicious node
which fits in the routes from the source node to destination node [8]. As soon as this
malicious node gets route request from the source node, it introduces itself as a node having
the shortest path to the destination by showing the minimum hop sum No. and maximum
sequence No. By performing this, the malicious node gets the chance to make it an active data
route element [9]. After this malicious node, capable of introducing wormhole attack in
WSNs by dropping all of the data Packets received from the source node [10]. Kim et al. [5]
have proposed a novel wormhole monitoring technique, also known as the algebraic
Watchdog. It allows nodes to monitor selfish behaviors probabilistically and also utilized
overheard Packets to regulate their neighbors locally. In this technique senders play an active
Machine Learning Techniques for Wormhole Attack Detection Techniques in Wireless Sensor
Networks
http://www.iaeme.com/IJMET/index.asp 339 [email protected]
role in the inspection of the node downstream. Baadache and Belmehdi [6] demonstrated a
novel approach to detect the wormhole attack by using an authentic end-to-end acceptance
based technique to evaluate the exact transmission of data Packets by transit nodes. However,
it has not considered the local information of nodes, therefore, unable to detect those
malicious nodes which have started dropping the Packets in currently.
Yang et al. [7] discussed an Anti-Wormhole Mechanism in order to detect the malicious
behavior of nodes. This technique utilizes wormhole monitoring nodes to detect the wormhole
attacker nodes. These nodes stay in sniff mode in order to monitor the mistrustful value of a
node depends upon the anomalous disparity among the communication Packets transferred
from the node. When a mistrustful value goes beyond a threshold, wormhole monitoring
node(s) will transmit a block Packet, informing other nodes to cooperatively isolate the
wormhole node. Poongodi and Karthikeyan [8] have proposed a method called Localized
Secure Architecture for WSNs. This method utilized security monitoring nodes which will be
activated if threshold value exceeded from the predefined value. If wormhole nodes are
monitored, then security monitoring nodes inform other nodes about the selfish node.
Banerjee et al. [9] proposed an AODV based wormhole attack mitigation technique in
WSNs without modifying the Packet format of AODV and without introducing any wormhole
monitoring Packets. Dasgupta et al. [10] provide a colored petri net model for monitoring and
prevention of wormhole attack in WSNs network. This model modifies a No. of properties
and provides better results as simulated through a CPN tool. Kurosawa et al. [11] in their
work provide a dynamic learning based technique for detecting wormhole attack in WSNs.
This technique is based on using dynamically updated training data for the isolating malicious
node. Jain et al. [12] make use of AODV’s sequence No. for mitigation of wormhole attack in
WSNs without modifying the Packet format of AODV. All the monitoring and prevention are
performed by an originating node without relying on other nodes in the network. Yong et al.
[13] makes use of neighbor set based along with the communication recovery technique for
mitigating wormhole attack in WSNs. Simulation results show that this technique reduced the
overhead of the network. Li et al. [14] in their work present a trust based on demand multipath
communication for isolating wormhole attack. A node’s trust is based upon its Packet
forwarding ratio. In this method, a source node creates numerous reliable paths to a
destination in solitary path discovery.
Namdeo et al. [15] provide an enhanced Watchdog based solution for protecting WSNs
against wormhole attack. In this, malicious node is detected on the basis of Packet flooding
and dropping parameters. Cai et al. [16] proposed a distributed wormhole monitoring system
for adversarial WSNs. This mechanism is used for preventing the networks from numerous
forms of wormhole attack. Packet delivery ratio is enhanced by using this distributed
approach. Imran et al. [17] provide monitoring and prevention technique for isolating
wormhole attack in WSNs. In this technique, DPS nodes are deployed in WSNs, that
uninterruptedly monitors the performance of their neighbor nodes. These DPS notice the
RREQs broadcasted by its neighbor nodes. After checking the No. of parameters of its
neighbor nodes, DPS node declares that suspicious node as the wormhole node and then
broadcast threat Packet on the network. Chatterjee et al. [18] in their work present a technique
for isolating wormhole attack in WSNs by using node stability system. This proposed
mechanism can successfully identify and isolate singular and co-operative wormhole nodes
from the network. Ghathwan et al. [19] introduce an artificial intelligence based technique for
preventing from the cooperative wormhole attack in WSNs. This is an integrated approach
based on both A* and Floyd-Warshall’s procedures. This mechanism works on the basis of
finding a shortest secure path for AODV (SSP-AODV).
Er. Harpal, Dr. Gaurav Tejpal and Dr. Sonal Sharma
http://www.iaeme.com/IJMET/index.asp 340 [email protected]
Babu et al. [20] discussed a novel honeypot based monitoring and isolation approach for
preventing from wormhole attack in WSNs. This proposed approach reduces the overhead,
Packet drop ratio and the communication load of the network. Kamatchi et al. [21] introduce a
new mechanism based on secret sharing and random multipath communication for preventing
from wormhole attack in WSNs. This Packet reduces the Packet delay and Packet drop ratio
in the network. Mohammed et al. [22] proposed a leader election based wormhole monitoring
system for mitigating wormhole attack in WSNs. For an optical leader election VCG model,
Cluster dependent and cluster independent concepts are used. Ritchie et al. [23] have
demonstrated a COB communication model by using complexity polynomial for preventing
against wormhole attack in WSNs. Performance of this technique is much better when
compared with Dynamic Source Communication. Chang et al. [24] proposed a Cooperative
bait monitoring scheme for protecting WSNs from collective attacks. This scheme works on
the basis of reverse tracing approach for modifying the performance of the network. Djenouri
et al. [25] have utilized Bayesian and social based techniques for mitigating malicious
attacker nodes in WSNs. This approach works on the basis of judgement to isolate the guilty
nodes from the network. Kaushik et al. [26] provide a solution for preventing the network
from both Wormhole and cooperative wormhole attacks. The drawback of modified AODV is
increased overhead.
Gong et al. [27] in their work presented a cooperative immune system for prevention of
WSNs against collective attacks. The concept of probability is used for analyzing and
detecting attacks. Ying et al. [28] discussed the threshold based wormhole monitoring system
for selective wormhole attack in WSNs. In this mechanism IDS nodes are set to sniff mode
for estimating the suspicious value of nodes. Arathy et al. [29] provide the Collaborative
Wormhole procedure for detecting single and collective attacks in WSNs. The proposed D-
MBH and D-CBH mechanisms are used for generating list of single, multiple and collective
wormhole attacker nodes. This approach reduces the computational overhead.
But, the review has shown that [1] - [29] have not focused on finding the link transmission
errors. The link transmission errors may occur in the WSNs because of the packets flooding
and due to maximum No. of nodes. Therefore, existing techniques have certain short coming,
which eventually decreases the WSNs performance. In order to handle this issue, a new
multipoint relay based Watchdog monitoring and prevention technique is proposed in this
paper. The proposed technique utilizes the dynamic threshold value to detect the wormhole
attacker node, and then clustering and Watchdog based optimistic path is selected for
communicating the Packets. Thus, it will reduce the overall Packet dropping, which will
improve the performance of the WSNs.
This paper is organized as: In section 2, the proposed wormhole monitoring technique is
discussed. In section 3, wormhole attack detection for different network layers is discussed.
The Simulation results of proposed technique using the MATLAB 2013a simulator are
discussed in section 4. The comparisons of the proposed technique with available state of the
art techniques are provided in section 5. In the last section conclusion and future directions
are also demonstrated.
2. PROBLEM STATEMENT
From the designed systems, the actual WSN AODV standard technique is adopted from [5].
The clustering based process is utilized in accordance with the good service quality, in which
every single node elects by itself along with the nodes in their transmission range. On
deciding upon cluster head, it is responsible to monitor the No. of multipoint relays (MPRs). It
determines the actual cost for every ith node by the following,
Machine Learning Techniques for Wormhole Attack Detection Techniques in Wireless Sensor
Networks
http://www.iaeme.com/IJMET/index.asp 341 [email protected]
wpl�a� = sc�a� × 1u�a�
���� (1)
Where sc�a� is actually the remainder of the mileage to get out of this path i.e. may be the
1-hop neighbor nodes in the similar route, Thus, the typical quickness from the ith node.
In comparison to [5] it is considered that the nodes for election simply for elected CHs of
which discuss its going direction. As soon as the attached CHs usually are determined, exactly
same solution is utilized in [5] to find the MPRs. As said before, wormhole strike reasons
package lower throughout the vast majority of the navigation techniques. From the put into
practice method, them goals MPR nodes, creating a considerable effect on multi-level
connectivity. For example, it is assumed that about 10% from the MPRs being malicious. It is
remarked that there were totally associated clustering with the direction-finding view prior to
assault took place, and everything the particular nodes can talk collectively easily.
Nevertheless, the moment vicious nodes are available; the particular amount of shut off
groupings retains escalating provided that quantity of nodes increases. This can be just
because which, if circle grows more packed, the particular nodes grow to be nearer together
plus associated by means of additional MPRs. For that reason, the quantity of opponents will
increase, which degrades the particular circle connectivity. This particular occasion illustrates
the necessity to establish a discovery process that may diagnose arsenic intoxication vicious
vehicles. Watchdog based mostly techniques usually are put in place inside related work [17].
The fact is that this kind of tracking tactics has got 2 key drawbacks. Very first, they
cannot separate whether or not some sort of small fortune falling function is a result of several
vicious assaults or merely legit causes, like small fortune collision. The 2nd challenge takes
place if Watchdogs have got troubles, although hearing panic or anxiety attack, as well as
accuses simple nodes to generally be misbehaving. To handle these issues, multipoint relay
based Watchdog monitoring and prevention technique is proposed in this paper. The proposed
technique will utilize the dynamic threshold value to detect the wormhole attacker node, and
then clustering and then Watchdog based optimistic path will be selected for communicating
the Packets. Thus, it reduces the overall Packet dropping, which improves the performance of
the WSNs.
In an effort to address the impact of the wormhole attack in WSNs, it demands to enhance
the performance of the well-known Watchdog monitoring scheme in tackling the collision
problem while monitoring. It should be noted that our main threat model assumes malicious
behavior from the MPR nodes during a Packet exchange among several nodes in the network.
The classical Watchdog technique fails to differentiate between collisions and attacks. A
remedy to this problem is the use of CL-layering, where different monitoring nodes from
different layers cooperates to enhance the performance of the single monitoring scheme.
Thanks to the cooperative and Cross Layer (CL)-layering features, any falsely reported
attacks can be handled. In this work, our main focus is to propose the cooperative CL layer
monitoring framework, where the Watchdogs are selected randomly. Finally, attacks prior to
Packet transmission specifically during broadcast transmissions can be detected utilizing the
Watchdog without the need to adopt any CL layering schemes as proposed in [17].
Er. Harpal, Dr. Gaurav Tejpal and Dr. Sonal Sharma
http://www.iaeme.com/IJMET/index.asp 342 [email protected]
3. MACHINE LEARNING BASED ATTACH WORMHOLE ATTACK
DETECTION
Table 1 Nomenclatur used
������ ����� �������
�� 10 � !! "!#$ℎ
& 2 () )* $ !!+
&,-_/$ 100 1,-/&2& /$! ,$/)�+
3))# 1 4))# +$, $/�5 * )& 1
61 2.1 Velocity factor 1
62 2.05 Velocity factor 2
-_#!,9 100 1,-/&2& ,�5! :,32!
;1, ;2 6 ;,�")& �2&=! +
>$!# 1: − 4)," ′,$$,B9_C+�. D **′ ",$, */3! /�$) $ℎ! &!&) E. >$!# 2: − D##3E 2�+2#! :/+!" */3$! /�5 )� $, 5!$ ,$$ /=2$! ,�" B)�:! $ /$
* )& �2&! /B $) �)&/�,3��; >$!# 3: − ()C +#3/$ )#$/)�+ , ! B)&! /� ,B$/)� $) */3$! $ℎ! ",$, /�$) 30
/70 *) &,$ /. !. 30 ",$, ,+ $ ,/�/�5 ,�" 70 ",$, *) $!+$/�5. >$!# 4: − ()C +!$ !K2/ !" B)�+$,�$+ >$!# 5: − L�/$/,3/M! ,�")& =!!+ ,�" 53)=,3 =!+$ ,�" 3)B,3 =!+$ :,32!+. N_O>P = ,�"���, &�; Q = ,�"���, &�; 3)B,3 =!+$ = ,�"���, &�; 53)=,3 =!+$ = ,�"�1, &�; >$!# 6: − ;!#!,$ $ℎ! +$!#+ 6 $) 10 Cℎ/3!�3))# <= &,-_/$� *) , = 1 $) �� *) = = 1 $) & ;1 = ,�"; ;2 = ,�"; >$!# 7: − T:,32,$! :!3)B/$E :_�$ = Q�/, U� + ;1 ∗ 61 ∗ �3)B,3 =!+$�/, U� − N_O>P�/, U�� + ;2 ∗ 62 ∗ �53)=,3 =!+$�1, U� − N_O>P�/, U��; >$!# 8: − 6ℎ!B9 =!+$ =!!+ 2+/�5 *)33)C/�5 B)�"/$/)�+: − /* :_�$ > :_#!,9 Q�,, =� = :_#!,9; !3+!/* :_�$ < −:_#!,9 Q�,, =� = −:_#!,9; !3+! Q�,, =� = :_�$; !�" N/�$! = N_O>P�,, =� + Q�,, =�; /* N/�$! > -_#!,9 N_O>P�/, U� = -_#!,9; !3+!/* N/�$! < −-_#!,9 N_O>P�/, U� = −-_#!,9; !3+! N_O>P�/, U� = N/�$! ; !�" >$!# 9: − ()C B,33 &!$, ,35) /$ℎ& /. !. >Q1�−O 100 − 4 − 1.7976931348623157T308 − [ 1.0 − \ 3.0 −P 1 − T 1 − > 1 − L 10 − ] ^!B/+/)�>$2&# ′� �; >$!# 10: − ()C 2#",$! >Q1 :,32!+ 2+/�5: − *) / = 1: �� # !+_N = */$_&�N_O>P�/, : ��; # !+_3)B,3 =!+$ = */$_&�3)B,3 =!+$�/, : ��; # !+_53)=,3 =!+$ = */$_&�53)=,3 =!+$�1, : ��;
Machine Learning Techniques for Wormhole Attack Detection Techniques in Wireless Sensor
Networks
http://www.iaeme.com/IJMET/index.asp 343 [email protected]
/* # !+_N < # !+_3)B,3 =!+$ 3)B,3 =!+$�/, : � = N_O>P�/, : �; !�" /* # !+_N < # !+_53)=,3 =!+$ 53)=,3 =!+$�1, : � = N_O>P�/, : �; !�" !�" ! �3))#� = # !+_53)=,3 =!+$; /* # !+_53)=,3 =!+$ <= 10^�−30� 3))# = &,-_/$; !3+! 3))# = 3))# + 1; !�" !�" O>P_/"- = B!/3�# !+_53)=,3 =!+$�; >$!# 11: − ()C "!*/�! ,$$ /=2$! $E#!+ ,+: − /. !. ′�2&! /B′, ′�)`abc ′, ′+$ /�5′, ′",$!′ ,�" ′ !3,$/)�,3′. >$!# 12: − ()C !:,32,$! �)`abc ,$$ /=2$! Qdef �)`abc3Qdef = B!33�1: �ghhi , O>P_/"-�; *) / = 1: �ghhi
/* +$ B&#/�,$$ /=2$!�E#!+{/}, ′�)`abc ′� �)`abc3Qdef{/} = , ,E*2��@�9� Bℎ, �^. ,$$ /=2$!�/ − 1�. Qdef�9 − 1��, 1: ^. ,$$ /=2$!�/ − 1�. �Qdef�; !�" >$!# 13: − ()C "!*/�! /�+$,�B!+ ",$, = 0; *) / = 1: �ghhi ",$,�: , /� = ^. ,$$ /=2$!�)D ,E�/ − O>P_/"-�; !�" >$!# 14: − B3,++m $!+$ abnh
�2&L�+$ = $!+$. �2&L�+$,�B!+��; #ifo = 0; #ifo Oipqn = 0; *) / = 1: �rbnh
# !"�/� = B3,++m /�+$ �$!+$ abnh�/ − 1� �;
#ifo Oipqn�/, : � = "/+$ /=2$/)�_*) _ /�+$ �$!+$ abnh�/ − 1� �; !�" >$!# 15: −()C B,33 ,�")& *) !+$ D35) /$ℎ&�B 3,+$ − 6 .9 − 1 2 ′� � >$!# 16: − B3,++m $!+$ abnh
�rbnh = $!+$. �rbnh��; #ifo = M! )+��rbnh , 1�; #ifo Oipqn = M! )+��rbnh , $icab_". �sdcnn���; *) / = 1: �rbnh
# !"�/� = B3,++m . B3,++m /�+$ � $!+$ abnh�/ − 1� �;
#ifo Oipqn�/, : � = B3,++m . "/+$ /=2$/)� t) L�+$,�B!� $!+$ abnh�/ − 1� �;
>$!# 17: − ()C !:,32,$! */�,3 )2$B)&!+. ;!$2 � */�,3 +)32$/)�
4. PERFORMANCE ANALYSIS
In order to assess the efficiency and competence of the proposed technique, i.e. proposed
technique and other some well-known wormhole monitoring techniques, MATLAB based
simulation is done for WSNs coding organizations and run wormhole monitoring and
prevention techniques. The existing and proposed wormhole monitoring techniques are
implemented on a Windows (2.4 GHz Intel i7 processor with 4 GB RAM and 1 TB memory).
The parameters used for simulation are shown in Table 1.
Er. Harpal, Dr. Gaurav Tejpal and Dr. Sonal Sharma
http://www.iaeme.com/IJMET/index.asp 344 [email protected]
Table 1 Simulation parameters
Parameter Value
Simulator used MATLAB 2013a
Simulation duration 4000 seconds
Area (meter) 100X100
No. of nodes 200
Communication technique AODV
Channel type Wireless
Packet size 4000 bytes
Mobility model Two ray ground propagation models
This section represents the comparison between some well-known wormhole attack
detection techniques with the proposed technique. The Average download efficiency is taken
as primary quality metric for comparison. It represents that how many packets are
successfully delivered within a given time. End-to-end delay is the mean time taken by a data
Packet to travel from source node to the destination node. This average time includes any type
of delay due to route discovery process along with a queue in data Packet transmission. In
this, only those Packets are included which are successfully transferred to the destination
node. This is calculated as:
ᵟ=u�vwx�
u�y� (22)
Where λ =Arrive Time, μ = Send Time, and η = No. of Connections.
Thmorer value of the Average download success rate is an indicator of the better
performance of the technique. Figure 8 shows the Average download success rate delay
comparisons of proposed technique with some existing approaches for preventing WSNs from
wormhole attack. It demonstrates that the proposed technique results in the increase in
average download success rate.
Figure 2 Average download success rate of proposed technique
The arrival time of the final packet is defined as the additional time taken to deliver
Packets at the destination. The arrival time of the final packet in the mobile WSNs network is
increased due to malicious node. The proposed approach results in decreasing the arrival time
Machine Learning Techniques for Wormhole Attack Detection Techniques in Wireless Sensor
Networks
http://www.iaeme.com/IJMET/index.asp 345 [email protected]
of the final packet of WSN as compare to existing procedures used for isolating the wormhole
attack as illustrated in Figure 3.
Figure 3 The arrival time of the final packet of proposed technique
Accuracy is the failure of transferring Packets to reach the destination. It happens due to
network congestion or some attacker node in the network. Accuracy is responsible for
reducing the Packet delivery ratio. It is calculated as:
Accuracy= δ-ρ (24)
δ= No. of Packets send from source and ρ= No. of Packets received at the destination.
Figure 4 Accuracy of proposed technique
Figure 4 shows the Accuracy comparisons of proposed technique approach with existing
procedures used for preventing WSNs. The figure clearly shows that the proposed technique
results in the decrease in Accuracy.
The Average download efficiency may increase if the attacker node is detected as early as
possible.
Er. Harpal, Dr. Gaurav Tejpal and Dr. Sonal Sharma
http://www.iaeme.com/IJMET/index.asp 346 [email protected]
Average download efficiency is defined as
Average download efficiency = u�
� (21)
Where ρ =No. of Packets received at the destination, ϒ=Simulation time
Figure 5 shows the Average download efficiency analysis of proposed technique with
some existing techniques. It depicts that the proposed technique after isolation of malicious
node results in the increase of Average download efficiency.
Figure 4 Average download efficiency comparison
5. CONCLUSION
The wormhole reduces the performance of the network a lot. Among the existing techniques,
Watchdog technique has better performance in preventing the wormhole attack. It utilizes the
local knowledge of the next hop node and eavesdrops it. In Watchdog technique, if the Packet
exchange time exceeds the threshold then node is marked as malicious. But, it has several
short comings; one of them is that it is unable to monitor link transmission error. In order to
handle this issue, a new multipoint relay based Watchdog monitoring and prevention
technique is proposed in this paper. The proposed technique utilizes the dynamic threshold
value to detect the wormhole attacker node, and then clustering and Watchdog based
optimistic path is selected for communicating the Packets. The proposed technique is
designed and implemented in the MATLAB 2013a tool. Comparisons have been drawn with
recently proposed techniques for monitoring and preventing against wormhole attack. The
performance analysis has clearly indicated that the proposed technique outperforms over the
available techniques. Thus, proposed technique has reduced the overall Packet dropping,
which improves the performance of the WSNs.
The authors declare no conflict of interest.
REFERENCES
[1] N. Marchang and R. Datta, Light-aggregation trust-based communication technique for
Wireless sensor networks, in IET Information Security, vol. 6, no. 2, pp. 77-83, 2012.
[2] E. Hernandez-Orallo, M. D. Serrat, J. C. Cano, C. T. Calafate and P. Manzoni, Improving
Selfish node Monitoring in WSNs Using a Collective Watchdog, in IEEE
Communications Letters, vol. 16, no. 5, pp. 642-645, 2012.
Machine Learning Techniques for Wormhole Attack Detection Techniques in Wireless Sensor
Networks
http://www.iaeme.com/IJMET/index.asp 347 [email protected]
[3] J. A. F. F. Dias, J. J. P. C. Rodrigues, F. Xia and C. X. Mavromoustakis, A Cooperative
Watchdog System to Detect Misbehavior nodes in Vehicular Delay-Tolerant Networks, in
IEEE Transactions on Industrial Electronics, vol. 62, no. 12, pp. 7929-7937, 2015.
[4] E. Hernandez-Orallo, M. D. S. Olmos, J. C. Cano, C. T. Calafate and P. Manzoni,
CoCoWa: A Collective Contact-Based Watchdog for Detecting Selfish nodes, in IEEE
Transactions on Mobile Computing, vol. 14, no. 6, pp. 1162-1175, 2015.
[5] M. Kim, M. Medard and J. Barros, Algebraic Watchdog: Mitigating Misbehavior in
Wireless Network Coding, in IEEE Journal on Selected Areas in Communications, vol.
29, no. 10, pp. 1916-1925, 2011.
[6] Abderrahmane Baadache and Ali Belmehdi, Struggling against simple and cooperative
wormhole attacks in multi-hop wireless WSNs networks, in Computer Networks, vol. 73,
pp. 173-184, 2014.
[7] Ming-Yang Su, Prevention of selective wormhole attacks on Wireless sensor networks
through wormhole monitoring systems, in Computer Communications, vol. 34, Issue 1,
pp. 107-117, 2011.
[8] Poongodi T., and M. Karthikeyan. Localized Secure Communication Architecture against
Cooperative Wormhole Attack in Wireless sensor networks, in Wireless Personal
Communications: pp. 1-12, 2016.
[9] Banerjee S., Sardar M. and Majumder K., AODV Based Black-Hole Attack Mitigation in
WSN, in Proceedings of the International Conference on Frontiers of Intelligent Computing:
Theory and Applications (FICTA). vol. 247,pp. 345-352, 2013.
[10] Dasgupta M., Santra D. and Choudhury S., Network Modeling of a Wormhole Prevention
mechanism in WSN, 4th IEEE International Conference on computational intelligence and
communication networks. pp. 734-738, 2012.
[11] Kurosawa S.; Nakayama H.; Kato N.; jamalipour A. and Nemoto Y., Detecting Wormhole
Attack on AODV-based Wireless sensor networks by Dynamic Learning Method, in
International Journal of Network Security. Vol. 5(3): 338-346, 2007.
[12] Jain S.; Jain M. and Kandwal H., Advanced procedure for monitoring and prevention of
cooperative black and Grayhole attacks in Wireless sensor networks, in International
journal of Computer Applications. Vol. 1(7), pp. 37–42, 2010.
[13] Sun B.; Guan Y.; Chen J. and Pooch U., Detecting Black-hole Attack in Wireless sensor
networks, in IEEE conference on Personal Mobile Communications, 5th European (Conf. Publ.
No. 492), 2003.
[14] Li, X.; Jia, Z.; Zhang, P.; Zhang, R. and Wang, H., Trust-based on-demand multipath
communication in Wireless sensor networks, in IEEE conference proceeding, IET
Information Security. Vol. 4, Issue 4, 2010.
[15] Namdeo M.; and Patheja P., Denial of Service (DoS) and Wormhole Attack Prevention by
Enhanced Watchdog Technique in WSN, in International Journal of Advanced Research
in Computer Science and Software Engineering. Vol. 5, Issue 12, 2015.
[16] Cai L.; Li, X. and Chong, P., A novel self-checking WSNs communication scheme against
active wormhole attacks, in Wiley online library, Security And Communication Networks.
Vol. 9, pp: 943–957, 2015.
[17] Muhammad I.; Khan A.; Haider A. and Mohsin I., Monitoring and Prevention of
Wormhole Attacks in Wireless sensor networks, in Springer, WSNs, Networks and Wireless.
Vol. 8629, pp 111-122.
[18] Chatterjee R. and Routray M., Wormhole Combat Using node Stability System in WSN,
in Social Informatics and Telecommunications Engineering. Vol. 62, pp 249-254.
[19] Ghathwan K. and Yaakub A., An Artificial Intelligence Technique for Prevent Wormhole
Attacks in WSN, Advances in Intelligent Systems and Computing. Springer International
Publishing Switzerland, 2014.
Er. Harpal, Dr. Gaurav Tejpal and Dr. Sonal Sharma
http://www.iaeme.com/IJMET/index.asp 348 [email protected]
[20] Rajesh M. and Usha G.,A Novel Honeypot Based Monitoring and Isolation Approach
(NHBADI) to Detect and Isolate Wormhole Attacks in WSN, in Wireless Personal
Communication. Springer, 2016.
[21] Kamatchi V.; Mukesh R. and Kumar R., Securing Data from Wormhole Attack Using
AODV Communication for Wireless sensor networks, in Advances in Computing&
Inform. Technology, Springer, AISC 177, pp. 365–373, 2013.
[22] Mohammed N.; Otrok H.; Wang L., Debbabi M. and Bhattacharya P., Mechanism Design-
Based Secure Leader Election Model for Wormhole Monitoring in WSN, in IEEE
transactions on dependable and secure computing. Vol. 8, no. 1, 2011.
[23] Ritchie L; Richa W. and Reisslein, M., Cluster Overlay Broadcast (COB): WSN
Communication with Complexity Polynomial in Source-Destination Distance, IEEE
transactions on mobile computing. Vol. 5, No. 6, 2006.
[24] Arathy K. and Sminesh C., A Novel Approach for Monitoring of Single and Collective
Wormhole Attacks in WSN, in Elsevier, Recent Advancements and Effectual Researches
in Engineering, Science and Technology – RAEREST. Vol. 25, pp. 264–271, 2016.
[25] Reena Jangra and Abhishek Bhatnagar. Comparison Analysis of Sensitivity of Noise B/W
Various Edge Detection Technique by Estimating Their PSNR Value. International
Journal of Computer Engineering and Technology, 6(10), 2015, pp. 01-12.
[26] Manpreet and Dr. Anil Kumar Sharma, Cryptographic Hash Key Algorithm to Mitigate
Wormhole Attacks and Lure Catch Algorithm to Block the Attackers, International
Journal of Computer Engineering and Technology, 7(3), 2016, pp. 108–117.
[27] Shailesh Birthariya, Sheetesh Sad Poornima Rawat, A Comprehensive Study of Signal
Detection Techniques for Spectrum Sensing in Cognitive Radio, International Journal of
Electronics and Communication Engineering & Technology (IJECET), Volume 5, Issue
11, November (2014), pp. 53-60
[28] A. EdwinRobert and Dr. M. Hemalatha, Behavioral and Performance Analysis Model for
Malware Detection Techniques, International Journal of Computer Engineering &
Technology (IJCET), Volume 4, Issue 1, January- February (2013), pp. 141-151
[29] Chang J., Defending against Collective Attacks by Malicious nodes in WSNs: A
Cooperative Bait Monitoring Approach, IEEE Systems Journal, pp.65-75, 2015.
[30] Djenouri D. and Badache N., Struggling against Selfishness and Wormhole Attacks in
WSNs, in Wireless Communications and Mobile Computing, UK: Wiley Online Library,
pp.689-704, 2008.
[31] Kaushik N. and Dureja A., Performance Evaluation of modified AODV against
Wormhole Attack in WSN, in European Scientific Journal, Portugal: University of the
Azores; .pp.182-193, 2013.
[32] Gong T. and Bhargava B., Immunizing Wireless sensor networks against collective
attacks using cooperative immune model, in Wiley Online Library, Issue: Security and
Communication Networks, pp.58-68, 2013.
[33] Ying S., Prevention of Selective Wormhole Attacks on WSNs through wormhole
monitoring systems, in Computer Communications (ELSEVIER), The Netherlands. pp. 107-
117, 2011.