With the popularity of multiplayer online games

iterative update cheats are emerging

seriously undermining game fairness. Direct memory access Direct Memory Access (DMA)[Figure 1] is a new type of cheating method that has emerged in recent years. Its basic principle is to use independent "special host" and DMA hardware to interact directly with the host computer's memory without going through the operating system or security monitoring program

thus bypassing the traditional anti-cheating detection mechanism. Specifically

cheat programs directly read or tamper with game running memory data through DMA channel to achieve perspective

self-aiming and other functions. Because data interaction is no longer restricted by conventional system permissions and protection measures

this kind of cheating has strong concealment and evasion ability

which brings higher challenges to existing anti-cheating systems. To solve this problem

this paper proposes a game anti-cheating method combining player behavior analysis and artificial intelligence model. We comprehensively utilize the player's operation timing data

telemetry information and game image features to learn the player behavior patterns by building deep learning models (such as CNN

LSTM

etc.) to identify abnormal operation trajectories and cheating patterns.The purpose of this study is to use novel anti-cheating methods

including(Aimbot)、(wallhack) and other common FPS game plug-ins have high detection accuracy

providing strong support for ensuring fair game environment.