Abstract
It is difficult to find spoofing traffic attack information for a wireless communication network, which leads to poor performance of spoofing traffic attack identification. Therefore, a spoofing traffic attack recognition algorithm for wireless communication networks based on improved machine learning has been designed. The process of network traffic classification and several common network cheating traffic attacks are analyzed. A chaotic algorithm is used to search and collect wireless communication network data, and Min-Max and z-score are used to standardize the collected data. The risk assessment function of wireless communication network spoofing traffic attack is constructed, and the spoofing traffic attack is preliminarily determined according to the function. The convolutional neural network in machine learning is improved, and the preliminary judgment results are input into the improved convolutional neural network to identify the attack behavior. The experimental results show that the recall rate of this method for wireless communication network spoofing traffic attacks can reach 90.08 % at the highest level, and the identification process takes only 1,763 ms at the lowest level. It can control the false positive rate of attacks below 4.68 % and the false positive rate below 2.00 %, and the identification effect of spoofing traffic attacks is good.