无线Mesh网络MAC性能的博弈分析及优化

Game-theoretic Performance Analysis and Optimization of WMNs

将博弈论用于IEEE 802.11无线Mesh网络媒体接入控制协议的性能分析和优化。通过将节点间的信道竞争过程建模为非完全信息动态博弈,求解博弈的纳什均衡,即各节点的最优分组发送概率,并据此提出改进的DCF协议(G-MAC):各节点首先通过监测信道,对当前博弈状态(竞争信道的节点数)进行估计;然后根据估计到的博弈状态调整其均衡策略(最小竞争窗口);最后通过有限次动态博弈获得最佳系统性能。同时,提出了一种能准确估计博弈状态的虚拟帧发送机制(VDCF),使空闲节点在转换为发送状态时可快速调整到均衡策略。仿真研究表明:G-MAC协议可以显著提高无线Mesh网络的系统吞吐量,降低延迟、延迟抖动和丢帧率。

A novel concept of incompletely cooperative dynamic game theory was used for performance analysis and optimization of IEEE 802.11 WMNs. The channel contention process was modeled as a dynamic game with incompletely information. According to Nash equilibrium of the game, a novel DCF scheme, called G-MAC, was proposed. Firstly, each node estimates the current state of the game （e. g. , number of competing nodes）. Secondly, each node changes its equilibrium strategy by tuning its local contention parameters （e. g. , the minimum contention window） based on the estimated game state. Finally, the game is repeated finitely to get the optimal performance. Besides, a virtual frame scheduling mechanism（VDCF）should be incorporated in G-MAC. Simulation results show G-MAC can increase the system throughput,decrease delay,jitter and packet-loss-rate.