Impact of Selfish Node Concentration in Manets

The communication in Mobile Ad hoc Network (MANET) is multi-hop in nature wherein each node relays data packets of other nodes thereby spending its resources such as battery power, CPU time and memory. In an ideal environment, each node in MANET is supposed to perform this community service truthfully. However this is not the case and existence of selfish nodes is a very common feature in MANETs. A selfish node is one that tries to utilize the network resources for its own profit but is reluctant to spend its own for others. If such behaviour prevails among large number of the nodes in the network, it may eventually lead to disruption of network. This paper studies the impact of selfish nodes concentration on the quality of service in MANETs.


INTRODUCTION
Mobile Ad hoc Networks don't rely on extraneous fixed infrastructure and can be installed without base station and dedicated routers.This makes them ideal candidate for rescue and emergency operations [1] and other short term networks.The nodes in these networks have limited battery power and bandwidth, and each node needs the assistance of others to get its packets forwarded.The conventional protocols in MANETs such as WRP [2], DSDV [3], AODV [4] and DSR [5] assume that all the nodes are cooperative and whenever a node receives a request to relay traffic, it always does so truthfully.
However the experience has shown [6,7,8,9] that as the time passes there is a tendency in the nodes in an ad hoc network to become selfish.The selfish nodes are not malicious but are reluctant to spend their resources such as CPU time, memory and battery power for others.The problem is especially critical when with the passage of time the nodes have little residual power and want to conserve it for their own purpose.Thus in MANET environment there is a strong motivation for a node to become selfish.
Marti et.al [7] have defined the characteristics of selfish nodes as follows: • Do not participate in routing process: A selfish node drops routing messages or it may modify the Route Request and Reply packets by changing TTL value to smallest possible value.
• Do not reply or send hello messages: A selfish node may not respond to hello messages, hence other nodes may not be able to detect its presence when they need it.
• Intentionally delay the RREQ packet: A selfish node may delay the RREQ packet up to the maximum upper limit time.It will certainly avoid itself from routing paths.
• Dropping of data packet: A selfish nodes may participate in routing messages but may not relay data packets The major reason for such behaviour is low residual battery power [6,8,9].It may here be clarified that a selfish node is not malicious and doesn't intend to involve itself in the network damaging activities such as content alteration, spoofing etc.It normally restrains itself from the activities of the other nodes which do not bring any benefit to it.
The literature provides [10,11,12,13,14] various strategies to deal with such behaviour.These strategies may categorise into two basic categories: Motivation/ incentive based approach and detect and exclude.
The motivation or incentive approach tries to motivate the users of the ad hoc network to actively participate in the forwarding activities.Such a system involves certain amount of money transfer to the relay nodes, on behalf of source or destination, to motivate them to forward messages [11,15,16].One of the motivation/ incentive based approach [8] is based on a virtual currency called nuglet.Every network node has an initial stock of nuglets.Either the source or the destination of each traffic connection use nuglets to pay the relay nodes for forwarding the traffic.The cost of a packet may depend on several parameters such as required total transmission power and the battery status of the intermediate nodes.Packets sent by or destined to nodes that do not have a sufficient amount of nuglets are discarded.The major drawback of this approach is the demand for trusted hardware to secure and maintain the record of the currency at central level.One such protocol, the ad hoc VCG [17] is based on the monetary transfer and discovers an energy-efficient path between the source and the destination.However, the number of messages that must be exchanged in order to find the route to the destination is quite high -in the order of O (n 3 ), where n is the number of network nodes.
Detect and exclude strategy avoids selfish nodes from the routing paths.This scheme uses two types of trust namely first hand trust and second hand trust.
• First hand trust: The node's personal observation about the neighbouring nodes.
• Second hand trust: The observation communicated by neighbouring node about the other neighbours of the network.
Watchdog and Pathrater [7] is a mechanism based on detect and exclude principle to deal with the selfish nodes.It uses Dynamic Source Routing [18] as base protocol.It has two components: Watchdog and Pathrater.The Watchdog is responsible for detecting selfish nodes that do not forward packets.The Pathrater assigns different rating to the nodes based upon the feedback that it receives from the Watchdog.Each node in the network buffers every transmitted packet for some time.During this interval, the node places its wireless interface into the promiscuous mode in order to overhear whether the next node has forwarded the packet or not and ratings are developed.These ratings are then used to select routes consisting of nodes with the highest forwarding rate.
SORI [22] is a protocol based on detect and exclude mechanism.It makes two record, the local evaluation record (First hand trust) and the overall evaluation record based on the reputation index given by the nodes about their neighbours.Each node in the network maintains tables of first and second hand trust of their neighbouring nodes.Based on these tables the trust of a node is calculated and then action is taken against the selfish nodes CONFIDANT [19] protocol adds trust manager and reputation index to the Watchdog and Pathrater mechanism.Each node in the network maintains two lists to deal with the selfish nodes.The nodes which behave rationally are kept in the friends list and the nodes which drop the packets or tamper them are kept in the black list.These lists are exchanged by the neighbouring nodes.Based on these list trust of a particular node is calculated.Whenever the trust value for a particular node falls below a certain threshold the protocol stops forwarding packets of that node.

THE CONCEPT AND THE ASSOCIATED WORK
All these strategies are generic and static do not take into consideration the concentration level of selfish nodes into network which may change dynamically.We are of the view that the strategy to deal with this selfish behaviour should be dependent on their concentration level in the network as the impact they have on the network activity will be different at different level of their concentration.
This paper studies the impact of selfish nodes concentration [0-100%] on the various Quality of Service (QoS) parameters [23, 24,25,26] related to performance of ad hoc network using MATLAB simulation.The QoS parameters taken into consideration are as follows: • Throughput: Percentage of packets received by the destination to the number of packets sent by the source.
• Hop count: Defined as the number of intermediate hops from source to destination.
• Packet dropped: Measure of the number of packets dropped by the routers due to various reasons.
• Probability of Reachability: Fraction of possible reachable routes to the all possible routes between all different sources to all different destinations.
The purpose of the proposed work is to help the various protocol designers to enable them in incorporating the dynamic strategies to deal with selfish nodes as their concentration varies.

SIMULATION AND RESULTS
A simulator was designed in MATLAB in which an area of 30 X 30 sq. unit's size was chosen.In total 40 nodes were distributed randomly in the given area, using randint function that uniformly distributes the random numbers, as shown in Figure 1.To create the route between a pair of random source and destination Dijkstra's shortest path algorithm was used.To avoid the selfish nodes in the path all the selfish nodes were made unreachable during that particular iteration.Figure 1.shows the snapshot of the simulation process wherein the red lines shows the shortest path when no node is selfish and the green lines shows the shortest path after the avoidance of selfish nodes.At low concentration of selfish nodes it is more likely green and red lines will be same and as the concentration increases they are likely to be different.Thus at 0% concentration of selfish nodes the route found between a pair of source and destination is shortest and at k% concentration the route found is shortest as if no selfish nodes were present.Thus with the increase in concentration of selfish nodes  Keeping these aspects in view and our simulator tried to make a quantitative assessment of the impact of selfish nodes on the above mentioned QoS parameters.This assessment will enable the designer of the protocols to combat the selfish node problem in the effective manner in an environment where the concentration of selfish node is changing dynamically.The quantitative assessment will enable the designer to make rational decisions while designing the protocol.

IMPACT ON AVERAGE HOP COUNT:
Figure 3. shows the impact of increase in average hop count as the concentration of selfish nodes increases.The average hop count is almost same when the selfish nodes are up to 10% of the total number of nodes.The average hop count increases to 2.5 times when the selfish nodes are nearly 100%.If the route formation does not occour, then in that the maximum hop count (16) was taken.There is no remarkable change in the percentage packet dropped when the selfish node concentration is up to 10%.It reaches to a maximum value of nearly 60% when the selfish node concentration is nearly 90%.

IMPACT ON THROUGHPUT
The results shows that there is no significant decrease in average throughput when the number of selfish nodes up to 10% of the total number of nodes.There is a generic trend of decrease in average throughput with the increase in concentration of selfish nodes.Since the process of source destination pair selection is random there are some fluctuations in the results as shown in Figure 5.It can be seen that the average throughput doesn't fall to zero even if all the nodes are selfish.The reason for this is that even at 100% selfish behaviour the communication between two immediate neighbours, as source and destination, still survives.

IMPACT ON PROBABILITY OF REACHABILITY
As expected the probability of reachability is nearly 100% when no node is selfish.As the number of selfish nodes increases the probability of reachability decreases but never reaches zero because even at 100% selfish behavior the communication between two immediate neighbors, as source and destination, still survives.Figure 6.shows the impact of selfish node concentration on probability of reachability.

CONCLUSIONS
The problem of selfish nodes is very common in ad hoc networks.The major reason for the selfishness is the loss of power with time.As the time passes away the nodes loose their battery power and in a disaster hit area or battle field area recharging may not be easily possible.The experimental results show that up to a concentration level of 10%, selfish nodes do not have remarkable negative effect on the network activities.As the concentration increases QoS becomes poorer and poorer though the network never comes to halt even if the selfish node concentration reaches to nearly 100%.The average hop count reaches to a maximum 2.5 times, Probability of Reachability and throughput comes down to nearly 50% at its peek and percentage of packet drop goes up to nearly 60% at the most.Since the concentration of selfish nodes increases with time, the routing protocols can be designed in such a manner that they can react to dynamic change in the concentration of selfish nodes.

Figure 1 .
Figure 1.Ad hoc Network • The average hop count is likely to increase • The packet drop rate is likely to increase • The average throughput is likely to decrease • The probability of reachability is likely to decrease

Figure 2 .
Figure 2. shows the flow chart for implementing the simulator on MATLAB.

Figure 3 .
Figure 3. Impact of Selfish Nodes on Average Hopcount

Figure 4 .
Figure 4.  shows the impact of selfish nodes concentration on the percentage of packets dropped.There is no remarkable change in the percentage packet dropped when the selfish node concentration is up to 10%.It reaches to a maximum value of nearly 60% when the selfish node concentration is nearly 90%.

Figure 4 .
Figure 4. Impact of Selfish Nodes on Average Packet Dropped

Figure 5 .
Figure 5. Impact of Selfish Nodes on Average Throughput

Figure 6 .
Figure 6.Impact of Selfish Nodes on Probability of Reachability [21]gned a intrusion detection system (IDS) for MANETs that consists of Local components: data collection, detection and response and Global components: cooperative detection and global response.Collaborative Reputation Mechanism (CORE)[21]is similar to the distributed IDS by Zhang et al. and consists of local observations that are combined and distributed to calculate a reputation value for each node.Based on this reputation, nodes are allowed to participate in the network or are excluded.In their work, the authors specify in detail how the different nodes should cooperate to combine the local reputation values to a global reputation and how they should react to negative reputations of nodes.An aspect that is not clearly stated in the work of Y. Zhang et.al.