光网络的光通路状态实时感知研究

Research on real-time perception of optical path status in optical network

为提高光通路状态感知的实时性,提出了一种光网络的光通路状态实时感知方法。首先对光通路态势要素进行提取,然后对提取出的态势要素进行量化处理,最后采用神经网络的方法对光网络的光通路状态进一步感知,以此完成光网络的光通路状态实时感知。实验结果表明,此次研究的光网络的光通路状态实时感知方法不仅减少了光通路状态获取时间,保持在2 min内,还减少了随机异常事件定位时间,仅为0.75 min,并提高了随机异常事件定位准确性,达到91%,满足光网络的光通路状态实时感知方法的设计需求。

In order to improve the real-time performance of optical channel state perception,a real-time perception method of optical channel state of optical network is designed. Firstly,the light path situation elements are extracted,and then the extracted situation elements are quantified. Finally,the neural network method is used to further perceive the light path state of the optical network,so as to complete the real-time perception of the light path state of the optical network. The experimental results show that the real-time sensing method of optical path state in optical network not only reduces the time of obtaining optical path state,which is kept within 2 min,but also reduces the time of locating random abnormal events,which is only 0. 75 min,and improves the accuracy of locating random abnormal events,reaching 91%,which meets the design requirements of real-time sensing method of optical path state in optical network.