Dual-modal system for real-time encryption and anomaly detection of 5G communication data integrating AES-GCM and LSTM

To address security and real-time challenges in 5G networks, this paper proposes a dual-modal system integrating AES-GCM encryption with an LSTM anomaly detection model, effectively balancing data confidentiality with anomaly identification.Experimentally, the system achieves optimal response speeds with encryption and decryption times of 18.4 ms and 21.7 ms, respectively.It demonstrates superior detection capabilities, attaining 98.1% accuracy -surpassing standalone models -while restricting missed detections to 1.0% and false positives to 0.5%.Additionally, the system excels in high-bandwidth environments, with encryption delays dropping significantly to 0.04 ms.In terms of resource efficiency, the hybrid model optimises CPU scheduling and lowers energy consumption to 130 J, compared to 165 J for traditional AES-GCM.Conclusively, this dual-modal architecture provides a robust, quantitatively validated solution that significantly enhances security, efficiency, and real-time processing in 5G scenarios.