A Novel Scalable Trust-Aware Deep Reinforcement Learning Algorithm for Energy-Efficient and Secure Routing in Software-Defined Wireless Sensor Networks for IoT

Abstract

Wireless Sensor Networks (WSNs) are the backbone of Internet of Things (IoT) ecosystems, but they remain constrained by limited energy, dynamic topologies, and increasing security threats. Conventional metaheuristic-based routing protocols typically optimize either energy efficiency or security, but rarely achieve both in a scalable manner. To address this research gap, we propose a trust-aware Software Defined Wireless Sensor Network (SDWSN) framework that integrates the Coati Optimization Algorithm (COA) for multi-objective routing with a hyperelliptic curve (HEC)-based blind signcryption scheme for lightweight yet robust data security. The novelty of this work lies in the joint optimization of energy, delay, trust, and hop count while simultaneously ensuring confidentiality, integrity, and anonymity through blind signcryption. Unlike traditional ECC and RSA, the proposed HEC-based scheme reduces computational complexity, making it suitable for resource-constrained IoT devices. The architecture leverages software-defined networking (SDN) programmability and the OpenFlow protocol to dynamically adapt routes based on real-time trust and energy metrics. Simulation results in NS-3 show that the proposed COA-HEC model significantly outperforms existing schemes (SEHR, IBFA, ESMR, GMPSO) by improving throughput (18.8 %–59.4 %), packet delivery ratio (by 4.8 %–12.4 %), and reducing average delay (up to 61 %) and energy consumption. The proposed framework establishes a scalable and secure routing paradigm for real-time IoT applications such as industrial automation, healthcare monitoring, and smart cities, thus advancing the state of the art in trust-aware SDWSNs.