一种优先级区分的机载无线网络MAC层自适应退避算法

An adaptive backoff algorithm in MAC layer for airborne network based on priority differentiation

针对现有无线网络媒介接入控制(MAC)协议退避机制存在的竞争窗口(CW)易产生振荡、不能为紧急业务提供低时延QoS保障、难以区分服务类别等不足,提出了一种区分业务优先级的自适应退避(PAB)算法。该算法在二进制指数退避(BEB)算法的基础上,根据信道忙闲标识和业务优先级自适应选择相应退避阶段,通过参数设置可对不同优先级业务相邻退避阶段的前、后转移概率进行调整。利用二维马尔可夫链理论对PAB进行了建模分析,理论分析获得了饱和吞吐量和平均MAC时延的数学表达式,仿真实验结果表明:在机载无线网络(AN)环境下,PAB的性能相对BEB有了显著提升,可有效保障机载无线网络多业务类型的传输需求。

The existing backoff mechanisms in the media access control（MAC）layer of wireless networks exhibit some shortcomings,e.g.,contention window（CW）oscillation occurs frequently,low delay QoS provisioning cannot be provided to ensure the demand of real-time traffic transmission,and different priority flows are not supported by different backoff schemes.To overcome the problems,a novel priority adaptive backoff（PAB）algorithm based on the binary exponential backoff（BEB）algorithm is proposed.In PAB,the backoff stage is chosen adaptively according to the channel busy identification and the services-priority,and the forwarding and receding transition probability between the consecutive backoff stages for different priority services can be adjusted for demands at any time.The 2DMarkov chain is introduced to model and analyze PAB and the analytical expressions for the saturation throughput and average MAC delay are derived and verified by simulations.The results show that in airborne network（AN）,the performance of PAB can offer a higher throughput and a lower delay than the BEB algorithm,and the transmission requirements of multi-service-types for ANs can be assured effectively.