AI-Enhanced Intrusion Detection for Gaming Security Using CBMA and Tree Physiology Optimization

Cyberattacks on gaming servers have grown to be a main security threat requiring effective intrusion detection systems. In this context, this work provides an AI-enhanced intrusion detection system for gaming security. In the proposed model tree physiology optimization (TPO) is used for feature section. On game servers, a Convolutional Block Multi-Attention (CBMA) model is implemented incorporating channel and spatial attention techniques for atteck classfication. With high AUC (0.9827) and only 900 trainable parameters, the suggested model is lightweight and efficient compared to GRU, LSTM, and RNN models, thereby guaranteeing cheap computational cost and rapid convergence and appropriate for real-time intrusion detection in gaming scenarios.