Study on automatic detection of abnormal traffic attacks in wireless communication networks based on edge computing
Zhaosheng Yang, Yan Feng Luo · International Journal of Internet Protocol Technology · 2024
In order to pinpoint the location of abnormal traffic attacks and reduce the miss rate of automatic attack detection, an automatic detection method for abnormal traffic attacks in wireless communication networks based on edge computing is proposed. Firstly, configure collection tools at each node to collect traffic data for normalisation, compensation, and filtering processing. Secondly, multi-dimensional features, such as time, spatial average traffic, skewness, kurtosis, etc., are extracted and fused, and a random detection sequence is constructed by accumulating mutation tendencies for abnormal traffic determination. Finally, edge computing based automatic detection of abnormal traffic attack location is set through distributed computing architecture, real-time analysis of network traffic, rapid location and determination of abnormal traffic attack location. Experimental results show that the maximum attack position detection error of our method does not exceed 1.0 m, and the maximum missed detection rate is only 1.09%, with a maximum decrease of 4.14% in missed detection rate.