03-04-2012, 01:07 PM
Modeling the Behavior of Selfish Forwarding Nodes to Stimulate Cooperation in MANET
[attachment=19402]
1. INTRODUCTION
Mobile Ad Hoc Network (MANET) is a collection of mobile nodes (hosts) which communicate
with each other via wireless links either directly or relying on other nodes as routers. The
operation of MANETs does not depend on pre-existing infrastructure or base stations. A mobile
node can become a failed node for many reasons, such as moving out of the transmission ranges
of its neighbors, exhausting battery power, malfunctioning in software or hardware, or even
leaving the network. Besides these failed nodes, based on the behavior, the mobile nodes are
classified into [3],[4],[5]:
• Cooperative Nodes are active in route discovery and packet forwarding, but not in
launching attacks
• Failed Nodes are not active in route discovery
• Malicious Nodes are active both in route discovery and launching attacks
Selfish Nodes are active in route discovery, but not in packet forwarding. They tend to drop data
packets of others to save their energy so that they could transmit more of their own packets and
also to reduce the latency of their packets. This type of attack comes under denial-of-service
(DoS) category.
EXISTING PROBLEM OF SELFISH NODE BEHAVIOR
In this section, we describe the problems caused by routing misbehavior of selfish nodes.
Selfish Node Problem
One immediate effect of node misbehaviors and failures in wireless ad hoc networks is the node
isolation problem due to the fact that communications between nodes are completely dependent
on routing and forwarding packets. In turn, the presence of selfish node is a direct cause for
node isolation and network partitioning, which further affects network survivability [6].
MODEL DESIGN AND ANALYSIS
Communications in an ad hoc network can be modeled as an infinitely repeated game. This kind
of models can describe situations in which the number of rounds is finite (as it happens in a
mobile ad hoc network, where nodes arrive, leave and move away changing neighborhood), but
there is not the knowledge on when the game is going to stop. Every node can not be sure that it
is going to play the next round with different opponents, since every node is moving.
Let us consider a “mobile” ad hoc network with two nodes that mutually need the other node to
reach (for example an access point) and that also exchange messages between them (Figure 2).
If there is a unique class, then there is not uncertainty about the type of the other node, and the
scenario is very simple. In the single shot scenario, Nash equilibria are (of course) dependent on
the value of á (and then on the energy class the nodes belong to).
CONCLUSION
We presented a model to describe behavior of an intermediary node in a route between a source
and a destination, which is a collaborative point of its one-hop neighbors. Our model is general
enough to describe cooperation enforcement mechanism that have been proposed in literature in
recent times, and it can be used to understand at what extent a node can be selfish, and how
much can we pretend from it.
From the investigations, it is found that model is able to regulate the selfishness based on
residual energy. With higher energy, the node is able to contribute more cooperation and as well
as more packet delivery ratio. Under steady state conditions, convergence of expected
cooperation depends on the number of neighbors in the cluster. More neighbors in the cluster
will bring more cooperation.
[attachment=19402]
1. INTRODUCTION
Mobile Ad Hoc Network (MANET) is a collection of mobile nodes (hosts) which communicate
with each other via wireless links either directly or relying on other nodes as routers. The
operation of MANETs does not depend on pre-existing infrastructure or base stations. A mobile
node can become a failed node for many reasons, such as moving out of the transmission ranges
of its neighbors, exhausting battery power, malfunctioning in software or hardware, or even
leaving the network. Besides these failed nodes, based on the behavior, the mobile nodes are
classified into [3],[4],[5]:
• Cooperative Nodes are active in route discovery and packet forwarding, but not in
launching attacks
• Failed Nodes are not active in route discovery
• Malicious Nodes are active both in route discovery and launching attacks
Selfish Nodes are active in route discovery, but not in packet forwarding. They tend to drop data
packets of others to save their energy so that they could transmit more of their own packets and
also to reduce the latency of their packets. This type of attack comes under denial-of-service
(DoS) category.
EXISTING PROBLEM OF SELFISH NODE BEHAVIOR
In this section, we describe the problems caused by routing misbehavior of selfish nodes.
Selfish Node Problem
One immediate effect of node misbehaviors and failures in wireless ad hoc networks is the node
isolation problem due to the fact that communications between nodes are completely dependent
on routing and forwarding packets. In turn, the presence of selfish node is a direct cause for
node isolation and network partitioning, which further affects network survivability [6].
MODEL DESIGN AND ANALYSIS
Communications in an ad hoc network can be modeled as an infinitely repeated game. This kind
of models can describe situations in which the number of rounds is finite (as it happens in a
mobile ad hoc network, where nodes arrive, leave and move away changing neighborhood), but
there is not the knowledge on when the game is going to stop. Every node can not be sure that it
is going to play the next round with different opponents, since every node is moving.
Let us consider a “mobile” ad hoc network with two nodes that mutually need the other node to
reach (for example an access point) and that also exchange messages between them (Figure 2).
If there is a unique class, then there is not uncertainty about the type of the other node, and the
scenario is very simple. In the single shot scenario, Nash equilibria are (of course) dependent on
the value of á (and then on the energy class the nodes belong to).
CONCLUSION
We presented a model to describe behavior of an intermediary node in a route between a source
and a destination, which is a collaborative point of its one-hop neighbors. Our model is general
enough to describe cooperation enforcement mechanism that have been proposed in literature in
recent times, and it can be used to understand at what extent a node can be selfish, and how
much can we pretend from it.
From the investigations, it is found that model is able to regulate the selfishness based on
residual energy. With higher energy, the node is able to contribute more cooperation and as well
as more packet delivery ratio. Under steady state conditions, convergence of expected
cooperation depends on the number of neighbors in the cluster. More neighbors in the cluster
will bring more cooperation.