一种基于统计的航空自组网多信道忙闲认知算法

传统自组网MAC协议采用载波侦听的方式判断信道忙闲,航空环境中节点分布广,通信距离远,网络拓扑动态变化,载波侦听结果并不准确,且侦听造成传输时延较大,信道利用率降低。结合AR预测,提出一种信道忙闲认知算法LS-AR(L-Step-Revise AR),通过统计过去若干时帧收到的突发个数预测当前时帧收到的突发个数,将多步预测值和真实值的差值作为下个时帧预测的修正,以预测突发个数判断信道忙闲,避免了采用侦听方式造成的时延,提高了信道利用率,为不同优先级分组接入信道提供了较为精确的判断依据。仿真结果显示,该算法对信道负载情况预测的准确率达到90%,满足航空自组网对MAC协议的要求。

A busy recognition algorithm of aeronautical ad hoc networks based on multi-channel Sstatistics

The traditional ad hoc network MAC protocol uses the carrier sense to judge the channel busy. However, the carrier sense results are not accurate because of the wide distribution of nodes, communication distance and dynamic network topology changes under aviation environment, and carrier sense causing large transmission delay, lower channel utilization. A channel busy cognitive algorithm LS-AR (L-Step-Revise AR) is presented combined with AR forecast, LS-AR predict the number of burst in current time frame according to the number of received burst in past several time frame, the difference between the predicted value and the true value as the amendment for the next prediction. And the predicted burst determine the channel busy, this way avoid delay caused by the use of carrier sense and improve the channel utilization, provides a more accurate judgment for the different priority packet access channel. Simulation results show that the algorithm can accurately predict the channel load conditions reached 90% and the meet the requirements for the MAC protocol of ad hoc networks.