INTERNATIONAL JOURNAL FOR TRENDS IN ENGINEERING & TECHNOLOGY

VOLUME 3 ISSUE 3 MARCH 2015 ISSN: 2349 9303

Using FRD Algorithm for Cluster Head

Selection and Certificate Revocation of
Adversary Nodes to Form a Secure Cluster in
MANET
V.Dhinesh1
1

Computer Science and Engineering

SVS College of Engineering
Coimbatore, India
dhineshhicet@gmail.com

M.Sivakumar2(Asist.Prof)
2

Computer Science and Engineering

SVS College of Engineering
Coimbatore, India
siva.recursion@gmail.com

AbstractThe major challenge in the wireless network service is to provide the guaranteed service. To overcome

this challenge we are going to use an important integral component called certificate revocation. In this paper, we
concentrate on the issue of certificate revocation to separate attackers from further contributing in the network
activities. By proposing the Enhanced Cluster based Certificate Revocation with Vindication Capability (CCRVC)
scheme, we can quickly and accurately revoke the certificates of the malicious nodes. We can improve the reliability
of the scheme by maintaining the Vote List (VL) by the Certificate Authority (CA). To enhance the accuracy, we
propose the threshold based mechanism to assess and vindicate the warned node is the malicious node or not. By
both numerical and simulation analysis, we evaluate the performance of our scheme. The demonstration result that
the proposed certificate revocation scheme is efficient and effective to guarantee secure communications in
MANET.
Index termsMANET, Certificate Authority (CA), Cluster Based Certificate Revocation with Vindication
Capability (CCRVC), Fuzzy Relevance Degreeand Security.

INTRODUCTION

ue to the mobility features such as ease of

deployment and dynamic topology, MANETs have
received increased attention in recent years. A lot
of increasing researches on the mobile distributed
computing. A MANET is consisting of number of nodes
with-out the infrastructure and those nodes are
connected with wireless communication. MANET is
self-organized wireless network, which consists of
mobile nodes that can freely move in the network such
as
laptops,
cell
phones,
Personal
Digital
Assistant(PDA), and the topology of the ad hoc network
isdynamical changing.These mobile devices are
cooperating with each other and forward the packets
between those nodes to extend the limited wireless
transmission range of each node by multi-hop relaying,
which is used for various applicationssuch as military
operation, disaster relief, emergency communications.
One of the most crucial requirements for network
service is security. The design of MANET is a major
challenge is to protect their vulnerabilities from security
attacks. Provisioning protected communications

IJTET2015

between mobile nodes in an antagonistic environment,

in which a malicious attacker can launching attacks to
disturb network security, is a primary concern. Owing to
the lack of infrastructure, mobile nodes in a MANET
have to implement all features of network functionality
themselves; they act as both end user and routers,
transmit packets for other nodes. MANET is the most
open network environment, here nodes can freely join
and leave from the network. Forcomplete security
solution, the MANET should encompass all the
threecomponents: prevention, detection, revocation.
Certification plays an important role in secure network
communication. The mobile nodes in the MANETs have
an additional functionality of forwarding to the other
nodes in the network. There is router to do the packet
forwarding operation in wired networks. The wireless
network is available for both legitimate networks, and
malicious attackers. Therefore, wireless on demand
MANETs are more susceptible to the attacks, while
comparing with wired. The certificate revocation
scheme provides fundamental security solutions such as
facilitating security, conveying real trust, ensuring
integrity and so on to all. The terrific amounts of

91

INTERNATIONAL JOURNAL FOR TRENDS IN ENGINEERING & TECHNOLOGY

VOLUME 3 ISSUE 3 MARCH 2015 ISSN: 2349 9303
research are happening on the certificate revocation
field, which includes the certificate distribution from a
third party. It ensures that each node in the network has
a trusted third party digital signature. Any node can
freely join in the network at a particular instant of time
for the attack detection. Certificate revocation is the
process of enlisting and removing the certificate of the
nodes which are launch attacks on the neighbor nodes.
In this paper, we are focusing primarily on the
security issue of MANET and also the process of
certificate revocation. I am going to use the different
method to select the cluster head and compare the
performance of the technique with the existing methods.
This paper organized as follows: In the next section
(section 2) we are going to give the brief overview of
certificate revocation techniques in MANET and also
analyze the advantage and disadvantage of voting based
and non-voting based scheme. Section 3 gives the
cluster head selection using theFuzzy Relevance Degree
algorithm. Section 4 gives the proposed schema used.
Finally, I conclude the paper in the following section.

2. RELATED WORKS
Researchers pay a much attention in ensures the
fundamental security of infrastructure less MANETs.
Due to the limited physical protection of nodes, the
dynamical topology changing, the vulnerability of
wireless link and the lack of infrastructure, it is difficult
to secure MANETs. In this section,
we
briefly
introduce the existing methods for certificate revocation.

2.1 Voting based mechanism

The certificate of the malicious nodes will be revoked
through the votes from the neighbor nodes. In this
schema, all nodes have to participate in the voting
system to evict the malicious nodes.
URSA proposed by Luo et al used a voting system to
evict nodes. In URSA the certificate authority issue the
valid certificate to the nodes, which are all join in the
network. The node with a valid certificate is considered
to be a legitimate node. A node before enter in to the
network, that will get valid certificate from the CA. The
major disadvantage of this voting based mechanism is, it
does not address the false accusation from their
neighbor malicious nodes.
Arboit et al in their work putsother mechanisms. It
allows the nodes in the network to vote together. The
nodes vote with variable weights is the major difference
from the Luo et al it ensuring the larger accuracy. By
using the reliability and trustworthiness of the node,
calculate the weight of anode, which derives from the
past behavior the node such as accusing other nodes and

IJTET2015

accusation from other node. When the weighted sum

from voters against the node exceeds a predefined
threshold, the certificate of the accused node will be
revoked. This is the way to improve the accuracy of the
certificate revocation.

2.2 Non-Voting based mechanism

In the non-voting based mechanism, any node with a
valid certificate can accuse a malicious node.
Clulow et al proposed a new strategy called suicide
for the common good which can quickly complete the
certificate revocation by only one accusation. In this
strategy the certificate of the accusing node will be
revoked by sacrificing itself to improve the accuracy of
the strategy. Due its suicidal strategy, the time taken to
revoke the certificate of the malicious node and
communications overhead of the certificate revocation
procedure is reduced. This suicidal strategy does not
concentrate on the false accusation so it degrade the
accuracy of the scheme
2.3 Cluster Based Certificate Revocation Scheme
Park et al proposed a new strategy called cluster
based certificate revocation scheme. The control
message is managed by the trusted third party and also it
maintains
Black list (BL) and Warned list (WL). The WL
contains accusing and accused node, the BL contains
malicious node. Any single neighboring node can
revoke the certificate of the malicious node. It also deals
with the issue of false accusation and it reduces the time
taken to revoke the certificate of the malicious node.
Wei Liu et al provide the vindication capability for
the certificate revocation technique. Before recovering
the certificate of the nodes in the WL, it verifies the
threshold value to vindicate the warned nodes as
legitimate node or not. In the existing system discussed
previously, a node proclaims itself as a cluster head. The
node periodically propagates CH Hello packet to its
neighbors. The method of CH selection is a faulty
method in the existing methods. So we are using the
Fuzzy relevance Degree to select a CH.
3. PROPOSED WORK
3.1Model of the Cluster Based Scheme
In this section, we present the cluster based
revocation scheme which can quickly revoke the
certificate of the malicious node. Clustering is a method
of grouping the nodes in the network based on some
criteria. I assumed that the cluster contain 10 to 50
nodes. The main aim is that the CH with in the cluster
has the rights to revoke the certificate of the malicious
node and also it maintains 2 different lists which contain
accusing and accused node with in the list called
Warned List (WL) and Black List (BL) respectively.
3.1.1 Cluster Construction
We proposed the cluster based architecture to construct

92

INTERNATIONAL JOURNAL FOR TRENDS IN ENGINEERING & TECHNOLOGY

VOLUME 3 ISSUE 3 MARCH 2015 ISSN: 2349 9303
the topology. Nodes in the network work together to
form the cluster consists of CH along with Cluster
Members (CMs) positioned with-in the transmission
range of the CH. Before the node joining in the network,
that should get a valid certificate from the third party,
which have the rights to distribute the certificate to the
nodes to communicate with each other freely in the
MANET.

3.1.2 Cluster Head Selection

By using a Fuzzy relevance Degree algorithm we can
maintain the cluster, to secure a cluster. This established
cluster allows the easier access and transmission of
packet from one secured cluster to the other cluster,
further happens reducing the transfer rate and also
reduces the effect of topological changes.
3.1.3 FSV Structure
The basic of the Fuzzy Relevance Degree algorithm is
FSV (Fuzzy State Viewing). This is a table structure like
algorithm showed in the figure. This table structure
algorithm consists of five parameters, which is
shownwith the Fig1. When a node communicates among
itself, the FSV comes in the transfer of packets. The
FRD value which is represented by will pass, when
two nodes communicate between each other. When two
nodes exchange the packets among each other, the range
of is form 0 to 1, which helps in the avoidance of
interference. By using the proposed FRD algorithm, we
are going to select the CH in the secured cluster. The
consequently result in the selection of Cluster Members.
The packet transmission among each node will be
through gateway node which helps in the broadcasting
of packets.
ID

Level
M-hop
Balance

M-hop(Multi-hop): The cluster creation and its

management is control is based on the value.
By separating the transfer process into multi
hops range from 1-hop to several multi-hops is
done by this parameter.
Balance: Each and every node in the cluster
should participate in the process of CH
selection. This parameter provides the way to
allocation of cluster head in the secure cluster.

Each node in the cluster must have the FRD value

denoted by . The FRD value is used to form a secure
cluster with high reliability and the optimum transfer of
data packets. This FRD value() is used by the nodes in
the communication process which is calculated by using
the parameter like mobility, distance and power. The
value of is range from 0 to 1. Cluster head is selected
from each node at once advocates itself. The energy of
the each node in the cluster must be stable and also the
characteristics of the with-in the cluster must be similar.
Figure 2 denotes the structure of the cluster before the
formation of the cluster. Figure 3 denotes the structure
of the cluster after the formation of the cluster. Any
node in the cluster, which having more power and signal
strength, derived from the RS will be selected as a CH.
In figure 3 C1,C2 clusters are communicating with-out
gateway node so it is not secure.

Fig 1.Packet Structure of FSV

The parameter of the packet is explained as follows

Identifier: Each and every node participates in
the cluster will have a unique identifier. It
helps to avoid the interference and make faster
the process of CH selection.
Fuzzy Relevance Degree(): the value of
Fuzzy Relevance Degree () is used to
communicate the nodes with each other. When
the node participating in the communication
process, the value will be sent by the
corresponding neighbor nodes. This value is
calculated by the parameters like mobility,
distance and power. The range of is from 0
to 1. Where the FRD value can be selected
accordingly, this will be formed as a set.
Level: This is the third parameter in the FSV
structure. The value is arranged into set
previously, this will be categorized into low
level, middle level and high level. The level
parameter is one of the criteria to select the
node as CH, CM and gateway node.

IJTET2015

Fig2. Structure before the cluster formation

Fig3. Modified Cluster structure

To balance the cluster formation as shown in the

figure3 C1, C2 is joined and act as a single cluster so
there is no need of any gateway node between C1 and
C2. By using the FRD value the above cluster formation

93

INTERNATIONAL JOURNAL FOR TRENDS IN ENGINEERING & TECHNOLOGY

VOLUME 3 ISSUE 3 MARCH 2015 ISSN: 2349 9303
was done. Acceptable values will be selected from the
set, which will support in the balancing of the structure
using the balance parameters. It also helps in
maintaining the process of allocating CH and
controlling the management process. Therefore, we
develop a secure cluster that has a high transfer rate in
terms of communication between nodes in the mobile
ad-hoc network.
3.1.4 Function of Cluster Head
The cluster contains a CH along with the CMs. The
main function of the CH is to form a secure cluster and
monitoring each node. The CH is in charge of updating
two list called Warned List (WL) and Black List (BL)
respectively. The BL contains the accused node which
was accused by more than the threshold value. The WL
contains both accusing and also accused node
respectively. The CH updates each list according to
receive the control packets. Each node can accuse a
node only once.

not, then the neighbor node will send the accusation

packet to the CH, which is shown in the Fig5(a). After
receiving the accusation packet from the neighbor node,
the CH should check the validity of the certificate of the
accusing node. The CH should verify that the voting
information in the VL which is maintained by the CA to
avoid false accusations. If it is valid the accusing and
accused nodes should enlist in the WL and number of
accusation against the same node is greater than the
threshold value then that will be listed in the black list.
Finally the revocation is broadcast to each node in the
cluster and any neighbor node can revoke the certificate
of the malicious node.

Fig5. Efficient Certificate Revocation

Fig6. Control packets

Fig4. System Architecture

3.1.5 Function of Certificate Authority

It is a trusted third party who is called as Certificate
Authority is developed in the cluster based scheme to
enable each node brings the certificate from it.
Certificate Authority is in charge of updating a list
called Vote List(VL). The CA updates the VL accords to
receive the control packets. CA broadcasts the
information of the VL to the entire network in order to
avoid the malicious node to accuse a legitimate node.
Each node with in the network should monitor the
neighbor node and vote for the node in the VL, which is
maintained by the CA.
3.1.6 Certificate Revocation
To revoke the certificate of the malicious node, we need
to consider 3 stages: accusing the node, verifying the
accusing node and notifying to each node in the cluster.
The revocation process is started by detecting the
malicious node. The neighbor node verifies the BL to
match whether the detected node is present or not. If

IJTET2015

3.2 List Management

3.2.1Warned List
It contains both accusing and also accused node. Any
node in the cluster can accuse a node if the certificate of
the accusing node is valid and it does not have negative
vote in the VL which is maintained by the CA. The
possible for the false accusation is very low so this
revocation scheme is highly reliable.
3.2.2 Black list
It contains only accused node which is also called as
malicious node. The BL will be broadcast to each CH to
revoke the certificate of the malicious node, which is
listed in the BL.
3.2.3 Vote List
The list, which is used to vindicate the accusing node, is
a legitimate or not. It contains four fields such as node
id, Cluster id, positive vote, negative vote.
ID
CID
Positive vote
Negative vote
Fig5. Vote List

ID: Each node in the cluster can have a

unique identifier for faster access and avoid
collusion.
CID: Each Cluster in the network should
have a unique cluster ID to differentiate the
clusters. 2 nodes can have the same ID, but
the CID of the 2 nodes should be different.

94

INTERNATIONAL JOURNAL FOR TRENDS IN ENGINEERING & TECHNOLOGY

VOLUME 3 ISSUE 3 MARCH 2015 ISSN: 2349 9303

Positive vote: Any node in the cluster should

be monitored by the neighbor node to check
whether the node is a malicious or not. The
node does not loss the packets and sends the
packets to the corresponding sink node the
neighbor node will give the positive vote.
Negative vote: If a node loss the packets and
it does not send the packets to the
corresponding sink node then the neighbor
node will give negative vote.

Whether a node has a single negative vote then that

corresponding node cannot accuse any other nodes in
the network.
4. CONCLUSION
In this paper, I have addressed a major issue to make
sure the security for MANET and also how to select the
CH for the secure cluster. The existing methods use the
method of neighbor sensing protocol to select the CH,
just by declaring the hello packets. But I proposed the
FRD method for the head selection process, including
various parameters like life time, battery power, distance
between them, velocity of nodes and also FRD value().
The revocation scheme is effectively done by the
neighboring node by exceeding the accusation against
the corresponding node. The false accusation is avoided
by maintaining the VL to verify the status of the
accusing node. Before the revocation process the status
of the accused node is compared with the VL. Therefore
the probability of occurrence of the false accusation is
very low. The efficient selection of CH and the process
of certificate revocation is help in the creation of a
secure cluster and also improve the performance of the
network. The network simulator is used to detect the
enhancement. The scheme increased the throughput by
the usage and it is efficient for so far.

REFERENCES
[1] H. Yang, H. Luo, F. Ye, S. Lu, and L. Zhang,
Security in Mobile Ad Hoc Networks: Challenges
and Solutions, IEEE Wireless Comm.,vol. 11, no.
1, pp. 38-47, Feb. 2004.
[2] INFOCOM 2004. Twenty-third AnnualJoint
Conference of the IEEE Computer and
Communications Societies (Volume: 1)
[3] L. Zhou, B. Cchneider, and R. Van Renesse,
COCA: A Secure Distributed Online Certification
Authority, ACM Trans. ComputerSystems, vol.
20, no. 4, pp. 329-368, Nov. 2002.
[4] P. Yi, Z. Dai, Y. Zhong, and S. Zhang, Resisting
Flooding Attacks in Ad Hoc Networks, Proc. Intl
Conf. Information Technology:Coding and
Computing, vol. 2, pp. 657-662, Apr. 2005
[5] B. Kannhavong, H. Nakayama, A. Jamalipour, Y.
Nemoto, and N Kato, A Survey of Routing

IJTET2015

Attacks
in
MANET,
IEEE
Wireless
Comm.Magazine, vol. 14, no. 5, pp. 85-91, Oct.
2007
[6] H. Nakayama, S. Kurosawa, A. Jamalipour, Y.
Nemoto, and N. Kato, A Dynamic Anomaly
Detection Scheme for Aodv-Based Mobile AdHoc
Networks, IEEE Trans. Vehicular Technology,
vol. 58, no. 5, pp. 2471-2481, June 2009.
[7] J. Newsome, E. Shi, D. Song, and A. Perrig, The
Sybil Attack in Sensor Network: Analysis &
Defenses, Proc. Third Intl Symp.Information
Processing in Sensor Networks, pp. 259-268, 2004.
[8] H. Luo, J. Kong, P. Zerfos, S. Lu, and L.
Zhang, URSA: Ubiquitous and Robust
Access Control for Mobile Ad HocNetworks,
IEEE/ACM Trans. Networking, vol. 12, no. 6,
pp. 1049-1063, Oct. 2004.
[9] G. Arboit, C. Crepeau, C.R. Davis, and M.
Maheswaran, A Localized Certificate
Revocation Scheme for Mobile Ad Hoc
Networks,Ad Hoc Network, vol. 6, no. 1, pp.
17-31, Jan. 2008
[10] J. Clulow and T. Moore, Suicide for the
Common Good: A New Strategy for
Credential Revocation in Self-organizing
Systems,ACMSIGOPS Operating Systems
Rev., vol. 40, no. 3, pp. 18-21, July 2006.
[11] K. Park, H. Nishiyama, N. Ansari, and N.
Kato, Certificate Revocation to Cope with
False
Accusations
in
Mobile
Ad
HocNetworks, Proc. IEEE 71st Vehicular
Technology Conf. (VTC 10), May 16-19,
2010.
[12] Wei Liu, Student Member, IEEE, Hiroki
Nishiyama, Member, IEEE, NirwanAnsari,
Fellow, IEEE, Jie Yang, and Nei Kato,
SeniorMember, IEEE , Cluster-Based
Certificate Revocation with Vindication
Capability for Mobile Ad Hoc Networks
[13] J. Lian, K. Naik, and G.B. Agnew, A
Framework for Evaluating the Performance of
Cluster Algorithms for Hierarchical Networks,IEEE/ACM Trans. Networking, vol.
15, no. 6, pp. 1478-1489, Dec. 2007.
[14] Dai Zhi-Feng,li Yuan-Xiang,HeGuo-Liang
Tong Ya-La Shen Xian-Jun,Uncertain Data
Management for Wireless Sensor Networks
UsingRough Set Theory,2006 IEEE.
[15] Ye
Tian,MinSheng,
Jiandong
Li,
YanZhang,Junliang Yao and Di Tang,Energy
aware Dynamic Topology Control Algorithm
for wireless AdHoc Networks,2008 IEEE

95