A BPNN-Based Traffic Anomaly Detection Method for Remote Managements of Optical Fiber Cores

With new network technologies and applications fast emerging in power telecommunications, network flow traffic exhibits many novel behaviors. However, to find abnormal part in network traffic is difficult. In this paper, we propose a new anomaly detection method to recognize abnormal network traffic for remote managements of optical fiber cores. Firstly, we use the Back Propagation Neural Network (BPNN) model to describe network traffic in these applications. Secondly, the BPNN training process is proposed to build network traffic model. The detailed detection method is proposed to find out the anomalous part in network traffic. Thirdly, the detection algorithm is proposed to recognize anomalous traffic. Simulation results shows that our approach is feasible and effective.