Cooperation in Wireless Networks with Unreliable Channels

Abstract

In a distributed wireless system, multiple network nodes behave cooperatively towards a common goal. An important challenge in such a scenario is to attain mutual cooperation. This paper provides a non-cooperative game theoretic solution to enforce cooperation in wireless networks in the presence of channel noise. We focus on one-hop information exchange and model the packet forwarding process as a hidden action game with imperfect private monitoring. We propose a state machine based strategy to reach Nash Equilibrium. The equilibrium is proved to be a sequential one with carefully designed system parameters. Furthermore, we extend our discussion to a general wireless network scenario by considering how cooperation can prevail over collusion using evolutionary game theory. The simulation results are provided to back our analysis. In particular, network throughput performance is measured with respect to parameters like channel loss probability, route hop count, and mobility. Results suggest that the performance due to our proposed strategy is in close agreement with that of unconditionally cooperative nodes. Simulation results also reveal how the convergence of cooperation enforcement is affected by initial population share and channel unreliability.