Adaptive trust-based secure routing protocol with reinforced anomaly detection for IoT networks

Abstract The increasing deployment of Internet of Things (IoT) networks has made them a target for sophisticated routing attacks, including blackhole, Sybil, rank, and wormhole attacks. To address these challenges, this paper proposes the adaptive trust-based secure routing protocol (ATBSRP), a novel framework that integrates trust evaluation, anomaly detection, and lightweight cryptographic mechanisms to enhance secure communication in IoT environments. The trust evaluation module dynamically assesses node behavior based on direct interactions, indirect recommendations, and historical trust scores, ensuring accurate and up-to-date trust values. The anomaly detection module uses a hybrid approach combining behavioral analysis, Gaussian mixture model (GMM)-based detection, and machine learning classifiers to effectively identify and mitigate malicious activities. Additionally, a lightweight cryptographic mechanism using elliptic curve cryptography (ECC), one-time hash chains, and a challenge-response mechanism safeguard data transmission with minimal computational overhead. The adaptive trust-based routing mechanism selects routes based on threshold-based trust scoring, ensures dynamic path adaptation in case of compromised nodes, and incorporates quality of service (QoS)-aware routing to maintain network efficiency. Experimental evaluations demonstrate that ATBSRP outperforms existing approaches in terms of packet delivery ratio (PDR ), end-to-end delay, throughput, routing overhead, and detection accuracy. The proposed framework offers a scalable, secure, and efficient solution for mitigating routing threats in IoT networks, ensuring reliable data transmission while minimizing network overhead.