Research on Network Traffic Anomaly Detection Algorithm

Network traffic anomaly detection is a difficult problem in network management. This paper proposed the wavelet generalized likelihood ratio WGLR) algorithm and the error performance detection (EPD) algorithm to solve this problem. The WGLR algorithm combines the generalized likelihood ratio (CLR) algorithm and the Wavelet transform method, and has an advantage of less computation cost. The EPD algorithm, which is based on the prediction error of the traffic model, will detect the anomalous change in the signal without test window delay. Simulating and network traffic experimental results show that the new algorithm has the better performance in network traffic anomaly detection.