Collaborative Intrusion Detection System to Identify Joint Attacks in Routing Protocol for Low-Power and Lossy Networks Routing Protocol on the Internet of Everything

Abstract

The Routing Protocol for Low-Power and Lossy Networks (RPL) routing protocol is utilized in the Internet of Everything (IoE) is highly vulnerable to various collaborative routing attacks.  This attack can highly degrade network performance through increased delay, energy consumption, and unreliable data exchange. This critical vulnerability necessitates a robust intrusion detection system. This study aims to enhance a Collaborative Intrusion Detection System (CIDS) for detecting and mitigating joint attacks in the RPL protocol, focusing on improving detection accuracy while minimizing network delay and energy usage. A series of algorithms and techniques are implemented, including Queue and Workload-Aware RPL (QWL-RPL) for congestion reduction, weighted random forward RPL with a genetic algorithm for load balancing, fuzzy logic for trust evaluation, and Light Gradient Boosting Machine (GBM) for attack detection. Additionally, Q-learning with a trickle-time algorithm is used to classify and manage joint attacks effectively. Numerical analysis indicates that the proposed approach performs better than existing methods in multiple metrics, including accuracy, energy consumption, throughput, control message overhead, precision, and computing time. By integrating these diverse techniques, the proposed CIDS offers a scalable and efficient solution to improve the security and performance of RPL-based networks in IoE environments, outperforming current approaches in detection accuracy and resource optimization.