A Hybrid Approach Using Spider Monkey and Tiered Particle Swarm Optimization for IoT Disturbance Recognition

A network of interconnected computing devices, sensors, and other electronic devices that can conversation and process information is known as the IoT. Devices on the network that have unique IP addresses may communicate with one another and with the outside world over the internet. The idea of the IoT has grown in significance because intelligent environments, including smart homes and cities, have been developed. These environments include a wide range of application disciplines. Aiming to make every day human activities more creative and competitive, such smart settings address concerns with living conditions, industrial demands, and energy usage. With the proliferation of IoT services and apps across many networks, the primary objective is now obvious. This study presented SMO-HPSO to improve identification accuracy while decreasing false alarm levels, and to lever the enormous quantity of intrusion statistics classification problems. In addition, the Random Forest Classifier (RFC) is used to find the best values for attack categorization using the UNSW NB-15 and NSL-KDD datasets. Current approaches such as RFC, DT, and SVM were contrasted with the suggested SMO-HPSO method. When tested against other algorithms, SMO-HPSO outperformed them on the NSL-KDD and UNSW NB-15 datasets. Positive outcomes, including enhanced efficiency in attack detection and categorization, have been shown by the suggested models.