In WiFi networks, mobile nodes compete for accessinga shared channel by means of a random access protocolcalled Distributed Coordination Function (DCF). Although thisprotocol is in principle fair, since it should guarantee that all thestations have the same probability to transmit on the channel,it has been shown that unfair behaviors may emerge in actualnetworking scenarios. These phenomena are due to different reasons,including non-standard configurations of the nodes, criticalnetwork topologies, and short-term performance observations.In this paper we propose a game-theoretic approach for definingan enhanced DCF scheme suitable for WiFi intelligent nodesemploying cognitive functionalities. We assume that a cognitiveWiFi node can dynamically change its strategy, by rationallytuning its contention window on the basis of channel observations.We prove that, for infrastructure networks with bidirectionaltraffic and homogeneous application requirements, our schemeis able to reach equilibrium conditions, which are also Paretooptimal. Specifically, we show that the station strategies convergetoward values which maximize a per-node utility function, whilemaintaining performance fairness.
|Number of pages||6|
|Publication status||Published - 2009|