A Behavior-Driven Lightweight Reinforcement Learning Framework for Secure Routing in IoT Networks

<title>Abstract</title> <p>The rapid growth of Internet of Things (IoT) networks has introduced significant routing security challenges, particularly under routing-layer attacks such as blackhole and selective forwarding. Traditional secure routing approaches rely on cryptographic mechanisms or trust dissemination, which introduce additional communication and computational overhead. This paper presents a Behavior-Driven Reinforcement Learning (BRL) framework for secure routing in multi-hop IoT networks. The proposed model integrates local forwarding behavior directly into the state and reward structure, enabling autonomous secure routing decisions without centralized control or explicit trust exchange. Simulation results demonstrate that BRL achieves up to 10–20% higher packet delivery ratio, lower routing overhead, improved energy preservation, and faster convergence compared to TBSIOP, RLBEEP, and LEACH. These findings confirm that behavior-aware learning enhances both security and efficiency in resource-constrained IoT environments.</p>