Channel-Change Games for Highly Interfering Spectrum-Agile Wireless Networks

The proliferation of wireless networks on unlicensed communication bands leads to coexisting networks, creating interference problems. In this paper, interference problems for spectrum-agile networks are addressed by allowing the networks to dynamically change channel. For insight into dynamic channel-change strategies, we model the networks as autonomous players in a multistage non-cooperative game-theoretic model. Here the networks are assumed to be highly interfering, i.e., when two or more networks exist on a single channel they cannot successfully carry traffic. Each network seeks to minimize its time to find a clear channel. The game-theoretic analysis reflects the motivations and choices of independent, rational, selfish decision makers that do not trust one another. We analyze the game-theoretic solutions appropriate for an untrusted environment, and compare the results with socially optimal decisions that would maximize the expected benefit of all coexisting networks in a trusted environment.