Research on interference identification technology in mobile communication networks based on GAN network

Due to the diverse nature of interference signals in mobile communication frequency bands and the difficulty in feature extraction, rapid and automated identification of unauthorized frequency use and malicious interference is challenging. To address this issue, this paper proposes an air interface interference identification technology for communication links based on air interface data collection. We construct an interference identification model to achieve automated interference identification. On the terminal side, by leveraging the advantages of signal feature extraction and deep learning, we extract time-frequency resource block (RB) occupation features and construct time-frequency RB feature maps. Using a semi-supervised Generative Adversarial Network (GAN), we learn the behavior of mobile communication time-frequency resource occupation under normal communication conditions. The trained model is then used to accurately identify interference signals, addressing the difficulties in interference signal feature extraction and automated identification. This method enables the detection of six types of interference on the terminal side under normal video communication activities. Experimental results show that the overall F1 score for identifying six types of interference using this method exceeds 0.9.