高速铁路光传送网络链路故障定位的研究

Research on Link Failure Location in High-speed Railway Optical Transport Network

铁路光传送网络OTN是列车安全运行的基础,由于OTN具有传输大带宽的能力,一旦网络出现故障将会造成传输业务中断,会严重影响列车的安全运行。因此,在故障发生时要快速、准确地定位故障位置,进而采取相应的故障恢复措施。光传送网络故障主要分为链路故障与设备故障。目前对于光传送网络链路故障定位是基于m-trail的故障定位方法,如何设置m-trail既能唯一确定链路是否出现故障,又能减少监测成本是研究的重点。本文针对光传送网络单链路故障定位提出基于RWS+MTA的m-trail分配算法。由于MTA算法设计m-trail的过程中在选择下一条链路时会固定选择最大权重的链路,因此会出现局部最优的情况。RWS+MTA算法采用一个概率模型可以在选择下一条链路时扩大搜索的空间,通过有限次迭代从中选出较优的分配方案,可以避免局部最优情况的出现。利用RWS+MTA算法设计了铁路骨干层二号环的m-trail分配方案,通过仿真验证了本算法在实现快速、精确定位的同时,可以有效降低运行时间以及监测代价。

Railway optical transport network （OTN） is the basis of the safe operation of trains. Because of the a bility of the OTN to transmit large bandwidth, the failure of the network will interrupt the transmission of service, which will seriously affect safe operation of the trains. Therefore, in the event of a failure, the location of the failure should be located quickly and accurately, and corresponding recovery measures should be taken. The failure of optical transport network is mainly divided into link failure and device failure. Currently, optical link failure location is based on m trail. The way to allocate m trail to uniquely determine whether the link is faulty and reduce the monitoring cost is the key point of research. In this paper, an RWS＋MTA based m trail allocation algorithm was proposed for single link fault location in optical transport network. Due to the selec tion of the next link in the process of the design of the m trail in the process of the MTA algorithm, the maxi mum weight link will be selected, which may result in a local optimal situation. The RWS＋MTA algorithm can expand the search space through using a probability model when selecting the next link, and can select the best allocation scheme through limited iterations, to avoid the emergence of local optimal situation. The RWS ＋MTA algorithm was used to design the m trail allocation scheme of the second ring of the railway backbone layer. The simulation results show that this algorithm can achieve fast and precise location, and effectively re duce the running time and the monitoring cost.