基于遗传算法的自适应CSMA/CA包调度

随着无线通信的发展，网络业务将逐渐由语音业务为主过渡到以不同QoS 要求的数据业务占主导地位, 简单的无线包调度算法如 max.C/I（最大信噪比）和WRR（公平轮循）
将无法满足不同数据业务QoS 需求。本文提出一种适应无线局域网特点的基于遗传算法的自适应包调度算法（AGAPS），该算法工作在数据链路层，这种算法适应无线网络；AGAPS根据信道状态、业务的QoS 需求、队列长度等参数来决定系统资源的分配，保护基站免受一些信息包流量的过度轮询，减少系统延时，提高系统的吞吐量。仿真结果表明AGAPS 算法对比max.C/I 算法有更好的用户公平性，而且系统吞吐量比WRR 算法高。
关键字：自适应包调度；QoS ；遗传算法
Abstract:With the development of wireless communications, network business will transit gradually from the Voice mainly to the different QoS requirements data-dominated business. Simple wireless packet scheduling algorithm as max.C/I and  WRR will not be able to meet the needs of different data QoS. A GA based adaptive packet scheduling algorithm (AGAPS) for WLAN was proposed in this paper, this algorithm worked in the data link layer, this algorithm adapt wireless networks. AGAPS according to the channel state, the QoS needs of business and the length of the queue to determine the allocation of the system resources. To protection of base stations from some of the packets flow over polling, reduce system latency and improve the throughput. Simulation results show that the proposed AGAPS have better user fairness than max .C/I and higher system throughput than WRR.
Keywords: adaptive packet scheduling；QoS； GA；