A Study On Nodes' Selfish Behavior In CSMA/CA Networks

With the increasing programmability of network adapters, selfish nodes can cheat for more bandwidth by controlling the contention window size in CSMA/CA networks, which will not only lead to the degradation of the network fairness, but also decrease the utilization ratio of the channel. Researchers often make unreasonable simplification to the model of the problem when applying game-theory approaches. As a result, these solutions are proved uneffective, because they lost the connection with reality.Based on the principles of Economics, we proposed an energy price mechanism solution. We adjust the allocation of bandwidth resources according to prices, which are assigned by extra energy consumption of each node when sending packets. In the energy price mechanism, each node can send a 'complaining' packet whenever it feels that it has been unfairly treated. To enhance the adaptability of the mechanism, we adopted a soft-state' approach.We adapted the mechanism design approach, in which the rational nodes will be included in our mechanism voluntarily. We suggest that the mechanism should be integrated in energy efficient routing protocols. In this way, the reservation utility of selfish nodes could be greatly decreased.In this paper, CSMA/CA network with selfish nodes is considered as a complex adaptation system, for the behavior of selfish nodes reflects the user's preference. Each node has its own proportion of the two perfect complements, namely throughput and survival time. Having implemented the energy price mechanism on Swarm simulation platform, we compared the effectiveness of several factors, such as the price of the resource, the 'soft-state' floating of price, the adaptive strategies of selfish nodes, etc. Via Swarm simulation, we testified that in CSMA/CA network, energy price mechanism solution not only ensured the fairness of channel bandwidth consumption among the nodes, but also improved the utilization ratio of the channel resources.