面向802.11b非对称无线网络容量优化的部分重叠信道分配方案

Partially Overlapped Channel Allocation Scheme for 802.11b Asymmetric Wireless Network Capacity Optimization

针对现有802.11b无线网络中部分重叠信道(POC)分配方案大多基于无向加权干扰图而忽略了实际网络中由于各链路传输功率和接收机灵敏度等因素导致的干扰不对称性问题,提出了一种有向加权图的非对称无线网络POC分配方案来优化802.11b非对称无线网络的全网容量。首先,构建有向加权干扰图,同时刻画信道距离、物理距离以及干扰不对称性;其次,将面向802.11b无线网络容量最优化的POC分配方案建模为局部互利博弈模型,网络中的每条链路是博弈的参与者,且效用函数设计为自身链路容量和潜在被干扰邻居链路的容量和;再次,证明所提博弈模型为势能博弈,存在至少一个纯策略纳什均衡点,同时最优纯策略纳什均衡为全网容量最优化的解;最后,设计了一个基于邻域合作的分布式多用户学习方案,实现了最优纳什均衡点的搜索。仿真表明,所提分布式学习方案可近似收敛到最优解,相比于现有其他算法可明显提高全网容量。

Most of the existing POC （Partially Overlapped Channel） allocation schemes in 802.11b wireless networks are based on undirected weighted interference graphs, which ignore the problem of interference asymmetry caused by factors such as transmission power of each link and receiver sensitivity in the actual network. Aiming at this situation, an asymmetric wireless network POC allocation scheme with directed weighted graph is proposed so as to optimize the whole network capacity of the 802.11b asymmetric wireless network. Firstly, a directed weighted interference graph is constructed while characterizing channel separation, physical distance, and interference asymmetry. Then, the POC allocation scheme optimized for 802.1 lb wireless network capacity is formulated as a local mutual benefit game model. Each link in the network is a participant in the game, and the utility function is designed as the capacity sum of its own link capacity and potentially disturbed neighbor links. Then, the proposed game model is proved to a potential energy game, and there is at least one PNE （Pure Nash Equilibrium）. At the same time, the optimal pure strategy Nash equilibrium is the optimal solution for the whole network capacity. Finally, a distributed multi-player learning scheme based on neighborhood cooperation is designed so as to realize the optimal Nash equilibrium point search. Simulation results indicate that this proposed learning scheme could converge to the optimal PNE and yield higher network capacity as compared with other existing schemes.