一种服务于流量特性的信道访问方法及公平性研究  

Study of a Channel Access Method Serving for Traffic Characteristics

作　　者：卢颖[1] 康凤举[2] 钟联炯[1] 

机构地区：[1]西安工业大学,西安710032 [2]西北工业大学,西安710072

出　　处：《小型微型计算机系统》2013年第7期1590-1594,共5页Journal of Chinese Computer Systems

基　　金：水下信息处理与控制国家重点实验室基金项目(9140C2305041001)资助

摘　　要：无线通信网络中,业务流在数学上呈现出一种统计上的自相似性.本文结合该流量特性,针对如何改善高负荷、高业务流量突发环境下网络性能的问题,给出一种基于节点突发流量的信道访问退避策略:通过动态监测和计算当前节点突发流量的大小来调整信道接入,从而减少碰撞次数,提高信道利用率;为解决算法公平性的问题,同时引入了节点流权值w(i).通过流量特性适应机制及节点流加权机制,使得网络在提高性能的同时能尽量减小算法不公平现象所造成的影响.理论分析及仿真结果均表明:在网络业务量突发性强、信道竞争激烈的情况下,该策略能有效缩短信道接入时延,排队长度等各项网络性能指标.In modern tactical communication network,traffic flow in math shows statistical self-similarity.This paper combining with this traffic characteristic raised a kind of channel accessing backoff strategies basing on nodes＇ burstness,in allusion to improve the netw ork performance issue under high load and high traffic burstness environment.It reduces collision times and increase use ration through dynamic monitoring and calculating current nodes＇ burstness.For solving the algorithm fairness issue,it introduces node-flow w eight coefficient w（i）.Through adaptive mechanism of traffic characteristic and w eighting mechanism of node-flow,it minimizes the influences of algorithm unfairness w hile enhancing performance.The results of theoretical analysis and simulation demonstrated： this strategy can reduce the channel access time-lag and queue length and improve relevant netw ork performance indexes effectively under strong netw ork traffic burstness and intense channel contention.

关 键 词：业务流量 自相似模型 竞争窗口 公平性指数 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]