Downlink Resource Auction in a Tree Topology Structured Wireless Mesh Network

We analyze the problem of downlink resource allocation in a non-cooperative multi-level tree topology structured wireless mesh network in which a selfish mesh router (MR) may refuse to relay other MRs' traffic so as to improve its own performance at the cost of overall system performance. Based on game theory, we propose an auction framework, where the parent MR serves as the auctioneer while its children MRs act as bidders and compete for time-slots. We derive a payment function from radio resource used for relaying traffic instead of money, so as to simplify the implementation and avoid the possible security problems from monetary payment. We prove the existence and uniqueness of Nash Equilibrium and propose a stochastic best response updating algorithm to allow the bids to iteratively converge to NE in a practical distributed fashion. Simulation results show the proposed auction algorithm greatly outperforms traditional algorithms in non-cooperative environments.