Integrating Long Short-Term Memory and Particle Swarm Optimization for Intrusion Detection in 5G Technologies

Extending coverage, decreasing latency, and increasing data rates are all requirements in the present situation for wireless communication. In response to these rising expectations, several new 5G technologies are being launched, including high-volume multiple-input multiple-output (MIMO) and device-to-device interaction. Adding more wireless components to the network, such as relays, small cell access points (SCAP), and hotspots, helps expand the coverage area. Yet, these parts are easily breached by intruders and serve as a gateway to be spread across the rest of the network. This work suggests combining the long short-term memory method with particle swarm optimization (LSTM-PSO) approach to strengthen intrusion detection system’s (IDS) security to overcome these shortcomings. Many concerns have been linked to IDS, including gradient vanishing, generalization, and overfitting. PSO and LSTM classification are used in the proposed system to address the gradient-clipping problem. The PSO is used to choose several effective features. The attack data is effectively classified and detected from normal data by applying the specified characteristics for effective classification inside an LSTM and PSO framework. The accuracy, precision, recall, and F1-score are used to evaluate the proposed LSTM-PSO framework’s performance. This study’s findings could benefit other areas, such as IoMT’s health monitoring system.