An Intrusion Detection System Using Machine Learning and Deep Learning Algorithms

The rapid development of IoT systems, IoT networks remain highly vulnerable to a wide range of security attacks, which can cause severe disruption to IoT services. To address these challenges, a novel intrusion detection system (IDS) is proposed using Message Queuing Telemetry Transport (MQTT) datasets. Initially, IoT data is collected from standard MQTT sources and processed in a pre-processing phase. The processed data is then used to extract three sets of features: the first set is optimally selected using the Improved Vulture Starvation–based African Vultures Optimization Algorithm (IVS-AVOA); the second set consists of statistical features refined by IVS-AVOA; and the third set is derived from an autoencoder optimized by the same algorithm. This proposed model effectively enhances attack detection in IoT environments.