Robust Deep Regressive Neural Learning Based Intrusion Detection and Prevention in Wireless Communication

Intrusion detection in 5G/6G networks is challenging due to their size. Traditional techniques fall short in handling the scale of these networks. To improve accuracy, a novel Robust Deep Regressive Neural Learning based Intrusion Detection and Prevention (RDRNL-IDP) Model is proposed. The model’s input layer processes training data, representing cellular node information. Hidden layers act as high-level feature extractors, predicting node features like Energy, Packet Loss Rate, and trust value. Leveraging these dynamic features, hidden layers distinguish normal and attack nodes. The output layer conducts robust regressive analysis, identifying relationships between normal node behaviors and independent variables like Dos, grayhole, or blackhole attacks. The RDRNL-IDP Model efficiently detects various intruders in 5G/6G networks with minimal time usage, providing improved accuracy and time complexity for intrusion detection and prevention during data communication. Simulation metrics include intrusion detection accuracy, detection time, and false positive rate, considering varying input cellular nodes.