Enhancing IoT Security Through Deep learning and Evolutionary Bio-Inspired Intrusion Detection in IoT systems

Advancements in systems based on the Internet of Things (IoT) have led to a significant transformation across various sectors. However, the security of IoT networks remains a major concern due to the diversity and ubiquity of connected devices. This paper introduces an innovative intrusion detection method for IoT systems, combining the bioinspired features selection algorithms with artificial neural network, emphasing a special focuse on Grey wolf optimisation algorithm (GWOA). Bio-inspired algorithms select the most relevant features from a dataset used in intrusion detection evaluation, while machine learning/deep learning (ML/DL) techniques ensure accurate classification of attacks. This approach provides an effective solution for enhancing IoT network security by identifying and responding to threats with precision and speed, thereby contributing to the protection of critical infrastructures against cyberattacks. The obtained results showed promising performances in intrusion detection with the optimal set of features. In which GWO achived a performance above 90% with approximately 20% of the global features set.