Optimized Intrusion Detection in Edge Computing Using Random Forest and Nature Inspired Algorithms
Varsha Sathasivam Banumathi, Swathi Sivakumar, Subasri Annadurai, Nithya Jayakumar, Ramya Jayakumar · 2025
Intrusion Detection Systems (IDSs) involving machine learning techniques is relevant to the mobile edge computing platform. Data traffic is booming and mobility increasing there, now. This paper discusses a really lightwight IDS for centralized infrastructure, by using nature-inspired techniques i.e. the ALO and ACO that max-out the performance of a system. These critical approaches are reflected in the data which thereby provides easy classifiers and detectors of mal/attacks. A deep, complex sophisticated machine learning approaches may be set aside while working with a relatively simplified yet effective method, the Random Forest algorithm - is capable of identifying and diagnosing harmful behavior yet reducing alarms without being alarmist. Our experiments show that this design improves the IDS in terms of speed and accuracy. It is an efficient approach to improve security without overloading resource-constrained devices in mobile edge environments.