Design and Optimization of Network Security Situation Awareness Algorithm for Generative Adversarial Networks Targeting Attack Data and Traffic

Network security situational awareness is crucial for detecting and mitigating network threats. However, the severe class imbalance in network attack data poses a significant challenge to accurate threat detection. To address this issue, this paper proposes a situational awareness algorithm based on Generative Adversarial Networks (GANs), focusing on three key aspects: attack data processing, traffic analysis, and perception platform development. Firstly, preprocess the UNSW-NB15 dataset and use conditional GAN to generate minority class samples to balance the dataset. Then, the Transformer model is used for precise attack behavior recognition, followed by hierarchical analysis for security assessment. The experimental results show that the algorithm is superior to existing oversampling methods, with an attack discrimination accuracy of 93.26% and a recall rate of 92.12%. Secondly, a security situational awareness algorithm optimized for real network traffic was introduced, which includes modules for traffic detection, data balancing, attack classification, and situational assessment. This algorithm significantly improves the accuracy of traffic analysis, with an accuracy rate of 94.65% and a recall rate of 95.01%. Finally, a network-based perception platform was developed, integrating functions such as account management, data analysis, and attack identification, providing real-time monitoring and situational analysis to improve the efficiency and reliability of network security management.