Reward based cooperative routing in Mobile Ad hoc Network by using Markov Decision Process

Mobile Ad hoc Network is a collection of mobile nodes, which communicate with each other in a wireless channel. In this Network, the nodes communicate in multi hop fashion with unconditional cooperation among the nodes. However in reality some nodes are selfish in their behavior, thus affecting the routing performance. Hence it is necessary to eliminate selfish nodes in the routing path to increase the routing performance. An alternative method to increase the routing performance is to encourage selfish nodes to become a cooperative node by providing appropriate incentives to selfish nodes. This paper discusses a methodology to provide incentives in the form as rewards. Rewards are formulated using Markov Decision Process. By using this methodology the cooperative node in the routing path expected to maximize the reward value for each decision epoch. The proposed Reward based AODV uses this reward value to establish cooperative route between source and destination node. The proposed routing algorithm is simulated and the simulation results prove that Reward based AODV routing algorithm is quite effective than AODV.