Detecting Selfish Node over the Active Path using Neighbor Analysis based Technique

Abstract

In this work, a nearest neighbor analysis has been performed to detect the selfish nodes in the active path and to generate a secure path. The existing AODV protocol is modified and a new bit is taken to define the trustful status. If status is 1, then node is valid, otherwise the node is selfish node and no communication is performed over that node. As the communication is performed, each node is analyzed by its neighboring nodes and builds a trust table. The reply status is 0 (by default) as the successful replied is received by a node, the value in the table changed to 1. Now the protocol checks the shared table and identifies the reply status. If the reply status is greater than the threshold value, the node is taken as the valid node and communication over that node is performed. This work has been implemented using NS-2.29 simulator and results shows that this technique is able to detect almost 90% selfish nodes in the active path.

Full text

International Journal of Science and Research (IJSR)
ISSN (Online): 2319-7064
Index Copernicus Value (2013): 6.14 | Impact Factor (2013): 4.438
Volume 4 Issue 3, March 2015
www.ijsr.net
Licensed Under Creative Commons Attribution CC BY
Detecting Selfish Node over the Active Path using
Neighbor Analysis based Technique
Sumiti1, Dr. Sumit Mittal2
1, 2MM University, MMICT & BM (MCA), Mullana (Ambala)
Abstract: In this work, a nearest neighbor analysis has been performed to detect the selfish nodes in the active path and to generate a
secure path. The existing AODV protocol is modified and a new bit is taken to define the trustful status. If status is 1, then node is valid,
otherwise the node is selfish node and no communication is performed over that node. As the communication is performed, each node
is analyzed by its neighboring nodes and builds a trust table. The reply status is 0 (by default) as the successful replied is received by a
node, the value in the table changed to 1. Now the protocol checks the shared table and identifies the reply status. If the reply status is
greater than the threshold value, the node is taken as the valid node and communication over that node is performed. This work has
been implemented using NS-2.29 simulator and results shows that this technique is able to detect almost 90% selfish nodes in the active
path.
Keywords: MANET, Selfish Node Attack, Nearest Neighbor Analysis
1. Introduction
Mobile ad hoc network (MANET) is a wireless network and
does not depend on the already existing infrastructure
because wireless nodes are capable of freely and dynamically
self-organizing in to the network and due to the reason
network topology changes fastly. In such networks, there are
different types of nodes and each node is capable of
communicating directly with any other node residing within
its transmission range. Mobile ad hoc network has few
problems like nodes moves randomly, dynamic network
topology and route changes; so there is a huge probability of
different types of attacks and packet losses. In the networks
different types of attacks are attacked by the attacker known
as malicious nodes and critical nodes. Malicious nodes are
those nodes that harmfully affect the network and other nodes
e.g selfish nodes. Attacks in Mobile and Ad hoc network
(MANET) can be classified as:
 Passive Attack
 Active Attack
In the passive attacks, the attacker only detects the data
which is exchanged in the network, and use it for their own
use without modify it or in other words only eavesdropping
of data. In a passive attack, the attacker's motive is just to
gather information about the network and the communication
pattern. But in the active attacks, the attacker modifies or
changes the data which is transmitted in a network. The
attacker has full control on the packet and can even inject or
drop the packet. In MANET attackers are also known as
comprised nodes. Malicious or compromised nodes are those
nodes in the network which are responsible for the active
attacks or damages the other nodes. Malicious nodes can
easily perform attacks by alter the information in the protocol
field, to destroy the transfer of the packets, to deny access
among the legal nodes.
In the MANET there are various types of attacks but we have
focused on the selfish node attack. Selfish nodes are those
nodes when the nodes receive the data and do not send to
next node; instead they use them to store their battery
lifespan, which they use for their own communications. The
efficiency of the network is greatly reduced by the selfish
nodes because they do not participate in the network
operations. Generally, it is easier for a node to become the
selfish node e.g. save resources for itself and ignore all
packets (data and control) that are not destined for it and
does not forward packets to next node. But some well-
behaved nodes in the network might not be required to
forward data packet. Examples of those scenarios are listed
as the following:
1) The node is located at the edge of the network. At that
location, the node does not have any other node to
forward data packet.
2) The network is already mature where all routing to every
possible destination has been established. A new node
then enters the network and wishes to use the network to
establish communication to another node. As long as
there is no link error, there is no change in the routing
table. The new node does not get any RREQ packet. As
a result, the new node does not be requires to do data
forwarding.
In this paper we have focused on detection of the selfish
nodes only active path. Below given figure shows about the
active and passive paths.
Figure 1: Shows the active and passive path
Paper ID: SUB152318
1295

International Journal of Science and Research (IJSR)
ISSN (Online): 2319-7064
Index Copernicus Value (2013): 6.14 | Impact Factor (2013): 4.438
Volume 4 Issue 3, March 2015
www.ijsr.net
Licensed Under Creative Commons Attribution CC BY
2. Related Work
The security problem and about the cooperation between the
nodes in network have been studied by different researcher in
the ad hoc network.
Watchdog and path rater [16] proposed an approach to detect
and isolate the misbehaving nodes. In this, a node forwarding
a packet checks if the next hop also forwards it. If not, a
failure count is incremented and the upstream node is rated to
be malicious if the count exceeds a certain threshold. The
path rater module then utilizes this knowledge to avoid it in
path selection. It improves the throughput of the network in
the presence of malicious nodes. However, it has the demerit
of not penalizing the malicious nodes.
Buchegger and Boudec[15] suggested that despite the fact
that networks only function properly if the participating
nodes cooperate in routing and forwarding. However, it may
be advantageous for individual nodes not to cooperate. They
propose a protocol, called CONFIDANT, which aims at
detecting and isolating misbehaving nodes, thus making
misbehavior unattractive. Here misbehaving nodes are
excluded from forwarding routes. It includes a trust manager
to evaluate the level of trust of alert reports. But it is not clear
how fast the trust level can be adjusted for compromised
node especially if it has a high trust level initially.
Singh et.al. [3] proposed the Detection and Prevention of
Blackhole Attack in Mobile Ad Hoc Network and quite
successful against black hole attack.
Chaba et.al. [4] proposed a Disable IP Broadcast Technique
for Prevention of Flooding-Based DDoS Attack in MANET.
Chaba et.al [5] have identified the mechanism for Detection
of Malicious Packet Dropping Based DDOS Attack in
MANET. However, none of these protocols are very
effective solution against PDDoS attacks in MANETs.
Poonam [8] proposed an opinion based cooperative trust
model to improve the performance of network, particularly in
the presence of malicious nodes. In the proposed model, each
node determined the trustworthiness of the other nodes with
respect to behavior observed. It calculated the direct trust by
the information obtained independently of other nodes and
indirect trust information obtained via opinion of other
nodes.
Poonam [9] proposed a novel method to enhance security in
both phases. Author presented the design of a routing
protocol based on trust, which ensures secure and
undisrupted delivery of transmitted data. An end to end
encryption technique has used to self encrypt the data without
the necessity of a cryptographic key.
Mobile Ad hoc networks are characterized by wireless
connectivity, continuous changing topology, distributed
operation and ease of deployment. The work proposed
contributes to detect the black hole attack using Modified
Associatively Based Routing protocol (MABR) which is the
modification and improvement of ABR [10].
Poonam [11] proposed a novel opinion based trust-aware
routing protocol (OBTRP) for MANETs to protect
forwarded packets from intermediary malicious nodes. In the
proposed model, each node determines the trustworthiness of
the other nodes with respect to behavior observed. It
calculated the direct trust by the information obtained
independently of other nodes and indirect trust information
obtained via opinion of other nodes.
3. Existing System
One of the biggest issues in mobile ad hoc network is that the
topologies change dynamically because of the node
movement. Not only this, even the nodes usually have no
predefined trust factor between each other. And another
biggest issue is the technical properties of regular nodes itself
like less energy of nodes. This property is most critical
because malicious nodes try to copy this and make it difficult
to distinguish between regular node and malicious node. A
malicious node from their behavior disturbs the network
operations and wastes the resources of regular nodes. But
intelligent malicious nodes elaborately choose a frequency at
which they cooperates the regular nodes.
4. Proposed Technique (Nearest Neighbor
Analysis using bit change based technique)
In this proposal, each monitoring node operates in
promiscuous mode and monitors both data and control
packets that are send around within its receiving range. Each
monitor node keeps record of its each neighboring node. In
this framework, a specific table is use for store the
information about the neighboring nodes. An extra field has
added in the table as the following:
1. Last Action
2. Last Request
3. Status
1. Last action of the neighbor node is that the last time seen
of neighbor node for contribution or providing services to
the network.
2. Last request of the neighbor node is that the last recorded
time of the neighbor node for last seen utilization or
requesting for services from the network. Monitoring
node updates these two fields every time, when take any
action for promiscuous mode.
3. Status is the current behavior of the neighboring node that
is detected by the monitoring node. The initial status for
any unknown node is set to zero and later on changed
according to their suspicious and behavior.
Whenever a monitoring node hears a request from its
neighboring node for forward a data packet, first it checks the
time difference between last request and last action of the
requestor. If it is within a threshold value means (TTL=1),
then called Action Hold off Value. If the value difference
exceeds the threshold, the status for the node has been set to
suspicious and for find out the status of the suspicious node a
Paper ID: SUB152318
1296

International Journal of Science and Research (IJSR)
ISSN (Online): 2319-7064
Index Copernicus Value (2013): 6.14 | Impact Factor (2013): 4.438
Volume 4 Issue 3, March 2015
www.ijsr.net
Licensed Under Creative Commons Attribution CC BY
special scenarios is used. In this testing, a fake RREQ packet
is broadcasted into the network. For minimize the traffic
flooding in the network, only the node that receives the data
forward request from the suspicious node is conducting this
test. In addition, this fake RREQ packet passes through one
hop (TTL=1). If it takes time more than (TTL=1) it means
this node is the selfish node and if it takes less than (TTL=1)
it means this node is the valid node. All monitoring nodes in
the neighborhood that detect this potential misbehavior and
waits for the suspicious node to rebroadcast the fake RREQ
packet within a certain timeout. If it responds the RREQ
packet, the status of the node is set to behave and the time of
its last action is updated. If it discards the packet and does
not respond, the monitoring nodes are labeled the suspicious
nodes as selfish.
5. Algorithm
Selfish_node_detection ( NAmax , NAi )
{
// NAmax= maximum value of average retransmission
numbers in the period //
if (NAmax – Nai < Threshold)
{
N k = non-selfish node;
}
else
N k = selfish node;
if(NAj= = 0)
{
N k is fully selfish node;
} }
Table 1: Simulation Parameter
Simulation Parameter
Value
Number of Nodes
50
Simulation Time
100 sec
Topology Size
700 X 700
Traffic Type
CBR (Constant Bit Rate)
Packet Size
512 bytes
Antenna Type
Omni directional Antenna
Routing Protocol
AODV
Queue Length Type
Drop Tail
Radio Propagation
Two Ray Ground
Total Packets
50
Channel Type
Wireless Channel
Network Interface Type
Wireless Phy
Performance evaluation is carried out using NS2. Nodes
moves according to the two Ray Ground mobility model.
And here aim is to implement AODV routing protocol for 50
nodes and sending CBR packets within random speed. First
the CBR files and scenario files are generated and then using
AODV protocol simulation is done which gives the NAM file
and trace file. The mobile node is simulated with a velocity
of 0-20m/s. It sends 300 CBR packets approximately. The
performance metrics are throughput and energy.
Attacker Nodes are shown in Red color example 4, 38, 39, 14
and these nodes drops packets from their nearest nodes.
Black blocks are shown in the screen shows the packet drops
from the normal nodes. Since there is no security to define so
packets are dropping during the route define because
attacker‘s nodes are present at nearest paths. For example:
node 4 is dropping packet from the node 36, node 38 is
dropping packet from the node 9 and node 39 from the node
40. For securely transfer data from source to destination node
without dropping packets we consider Node 1 is server. And
this server (1) checks communication within each node for
securely data transfer without using computational
assumption.
Figure 2: Normal simulated nodes
Figure 3: Selfish nodes dropping packets
Figure 4: Showing secure route
According to the figure 5, Source node 0 sends data to Server
Node 1. Black dot is shown between 0 and 1. After this data
communication checking is done between the black node 48
and 13. Then data communication checking is done between
the black node 13 and 45. After communicate with these
nodes in this path, a new routing path is selected from source
to destination. Blue color nodes shows a new secure path for
data transfer from source to destination node example 0—2--
-26—7—22---48 – 13 – 45 – 44. By using this path data is
securely transfer without packet loss.
Figure: 5 Screenshot shows secure data transfer through the
specified path without packet loss.
Paper ID: SUB152318
1297

International Journal of Science and Research (IJSR)
ISSN (Online): 2319-7064
Index Copernicus Value (2013): 6.14 | Impact Factor (2013): 4.438
Volume 4 Issue 3, March 2015
www.ijsr.net
Licensed Under Creative Commons Attribution CC BY
Below given figure shows the comparison of Packet Delivery
Ratio (PDR) for Selfish AODV (four nodes acts as selfish)
and Normal AODV (No selfish node). Graph for Normal
AODV shows that there is no any loss of data packets means
no one node acts as a selfish node ie. 100% PDR and But
graph for Selfish AODV shows that there is loss of some
packets, all packets are not delivered. In this, some nodes act
as selfish nodes. And these nodes find out with the help of
our neighboring node based system for MANET to detect
selfish node.
Figure 6: Shows the result
6. Conclusions
Mobile ad hoc network is a famous upcoming field of
research with practical applications. Because there are many
reasons of attacks like lack of regular security infrastructures,
changing network topology and open medium of
communication, which are not totally secured. The analysis
shows the real challenge of this study is that the throughput
performance depends on the number of paths in the network
and the locations of malicious nodes. And the communication
checking between nodes depends on the detection range (that
a server is in the range of a sender and a malicious node).
Simulation results shows the effect of network sizes, numbers
of nodes and mobility speed that helps to understand the
impact of the packet dropping attack and its mitigation.
References
[1] R.Gopal, V.Parthasarathy, A.Mani, ―Techniques to Identify
and Eliminate Malicious Nodes in Cooperative Wireless
Networks‖, IEEE International Conference on Computer
Communication and Informatics (ICCCI -2013), Jan 2013.
[2] Dipali Koshti and Supriya Kamoji, ―Comparative study of
Techniques used for Detection of Selfish Nodes in Mobile Ad
hoc Networks‖ International Journal of Soft Computing and
Engineering (IJSCE) , Volume-1, Issue-4, September 2011.
[3] Yudhvir Singh, Yogesh Chaba, ―Detection and Prevention of
Blackhole Attack in Mobile Ad Hoc Network‖, In proceeding
of IEEE International Conference on Advance Computing
Conference, IACC 2009; March 6-7, 2009; pp 2668-2672.
[4] Yogesh Chaba, Yudhvir Singh, Preeti, ―Performance Analysis
of Disable IP Broadcast Technique for Prevention of Flooding-
Based DDoS Attack in MANET‖, Journal of Networks,
Academy Publisher, ISSN: 1796-2056.
[5] Yogesh Chaba, Yudhvir Singh, Preeti, ―Detection of Malicious
Packet Dropping Based DDOS Attack In MANET‖, Journal of
Computer Science, ISSN 0973-2926; Vol. 3, Issue 2, Jan-Feb
2009; pp 959-964.
[6] Yogesh Chaba, Yudhvir Singh, Preeti, Deepak, ―Performance
analysis of various distributed denial of service based attacks
in mobile adhoc networks‖, IEEE International Conference on
Advance Computing Conference, IACC 2009; March 6-7,
2009; pp 3228-3132.
[7] Preeti, Yogesh Chaba, Yudhvir Singh, ―Review of Detection
and Prevention Policies for DDoS attacks in MANETs‖, In
Proceedings of 2nd National Conference on Challenges &
Opportunities in Information Technology (COIT-2008) RIMT-
IET, Mandi Gobindgarh; March 29, 2008; pp 56-59.
[8] Poonam, "Eliminating misbehaving nodes by Opinion based
Trust Evaluation Model in MANET‘s‖, ICCCS‘2011,
Rourkela, Odisha, India, ACM 978-1-4503-0464-1/11/02.
[9] Poonam Gera, "Trust Based Multi-Path Routing for End to
End Secure Data Delivery in MANET‘s", SIN‘2010,
Taganrog, Rostov-on-Don, Russian Federation, ACM 978-1-
4503-0234-0/10/09.
[10] M.Shobana, "Geographic Routing used in MANET for Black
hole Detection", CCSEIT-2012, Coimbatore, Tamilnadu,
India, ACM 978-1-4503-1310-0/12/10.
[11] Poonam, "Misbehaving nodes Detection through Opinion
based Trust Evaluation Model in MANETs", International
Conference and Workshop on Emerging Trends in Technology
(ICWET 2011), TCET, Mumbai, India, ACM 978-1-4503-
0449-8/11/02.
[12] Preeti, Yogesh Chaba, Yudhvir Singh, ―Distributed Denial of
Service: Taxonomies of attacks and Countermeasures‖, In
National Seminar on ‗Enterprise Information Systems‘ at
Apeejay College of Management, Jalandhar; May 24, 2008
[13] Y. an Huang and W. Lee, ―Attack analysis and detection for ad
hoc routing protocols‖, in Proceedings of The 7th International
Symposium on Recent Advances in Intrusion Detection (RAID
2004), September 2004.
[14] S. Buchegger and J.-Y. L. Boudec, ―Performance analysis of
the confidant protocol (cooperation of nodes: Fairness in
dynamic ad-hoc networks)," in MOBIHOC'02, 2002.
[15] J. Mundinger and J.-Y. L. Boudec, ―Analysis of a reputation
system for mobile ad-hoc networks with liars," wiopt, vol. 00,
pp. 41{46, 2005.
[16] L. Buttyan, and J. Hubaux, ―Enforcing cooperation in self
organizing mobile as hoc networks,‖ In Proceedings of
IEEE/ACM Workshop on Mobile Ads Hoc Networks,
Technical reportDSC/2001/046, EPFL-DIICA, August 2002.
[17] F. Kargl, A. Klenk, S. Schlott, and M.Weber, ―Advanced
detection of selfish or malicious nodes in ad hoc networks‖,
August 2004.
[18] Radhika Saini and Manju Khari, ―Defining Malicious
Behavior of a Node and its Defensive Methods in Ad Hoc
Network‖ International Journal of Computer Applications,
Volume 20, No.4, April 2011.
[19] Jaswinder Singh and Ramandeep Kaur, ―Towards Security
against Malicious Node Attack in Mobile Ad Hoc Network‖,
IJARCSSE Volume 3, Issue 7, July 2013.
[20] S. Marti, T. Giuli, K. Lai, and M. Baker, ―Mitigating routing
misbehavior in mobile ad hoc networks‖ in The 6th ACM
International Conference on Mobile Computing and
Networking, 2000.
[21] Prabha Rani, Yogesh Chaba, Yudhvir Singh, ―Hybrid
Approach for Detection and Prevention of Blackhole Attack in
Mobile Adhoc Network‖, International Journal of Wireless
Communication, ISSN 0974-9640, pp 885-890, (August,
2011).
Paper ID: SUB152318
1298