Cooperative and Non-Cooperative Control in IEEE 802.11 WLANs
Résumé
Numerous algorithms and techniques for optimal performance of an IEEE 802.11 WLAN have been investigated by researchers. These algorithms make use of either power control or PHY (physical layer) rate control (i.e., adaptive selection of PHY rates) or both to achieve maximum throughput levels for the network at minimum power consumption by the mobile devices. However most of these techniques are non-cooperative by definition (i.e., they attempt to maximize an individual node's performance and not the overall network performance). In this report, we analyse cooperative and non-cooperative rate and power control based on an expression for the throughput of a node in an 802.11 WLAN that uses the Distributed Coordination Function (DCF) with an RTS/CTS frame exchange. We formulate a payoff function comprising of the throughput and costs related to power consumption of a mobile node. The payoff function is optimized and closed form expressions for the optimal PHY rate are obtained. In the cooperative approach we seek to obtain the optimal rates under two different scenarios - max-min fair rate and global multirate allocation. In the non-cooperative game approach we consider only multirate allocation. Our main contribution is that we obtain explicit expressions for the optimal PHY rate which in turn can be used to compute explicit expressions for the throughput. We consider optimization problems for both finite number of nodes n and for the limit $n \to\infty$. Single node throughputs corresponding tothe optimal PHY rates are numerically studied and it is observed that network performance in the cooperative scenario is superior than in the non-cooperative scenario.
Loading...