NeuralGuard5G: Enhancing 5G Network Slice Security Against Distributed Denial of Service Attacks

Network slicing, a fundamental aspect of fifth-generation (5G) cellular networks, presents unparalleled flexibility and customization but grapples with significant security challenges, notably susceptibility to DDoS attacks. This work introduces NeuralGuard5G is an innovative system leveraging LSTM deep learning techniques. NeuralGuard5G autonomously distinguishes User Equipment (UE) network traffic as normal or indicative of a DDoS attack, enhancing detection accuracy and dynamically assigning an appropriate network slice to legitimate UE requests. Notably, the framework eliminates the need for predefined thresholds and minimizes computational burdens related to feature engineering. The analysis thoroughly examines feature selection methods, utilizing Random Forest (RF) to optimize feature selection and enhance the depth of DDoS attack detection. Empirical evidence supports NeuralGuard5G's efficiency, adaptability, and scalability, positioning it as a robust solution for fortifying network slices against DDoS threats in the evolving landscape of 5G networks. This comprehensive approach contributes to preserving the overall integrity and availability of resources, making NeuralGuard5G a valuable asset in enhancing the security posture of 5G networks.