Priyadarshi Rahul, Kumar Ravi Ranjan, Ranjan Rakesh, Kumar Padarti Vijaya
Faculty of Engineering and Technology, ITER, Siksha 'O' Anusandhan (Deemed to be University), Bhubaneswar, 751030, India.
Department of ECE, National Institute of Technology, Patna, Bihar, 800005, India.
Sci Rep. 2025 Jul 1;15(1):22292. doi: 10.1038/s41598-025-08677-w.
This paper proposes a modular Artificial Intelligence (AI)-based routing framework for Wireless Sensor Networks (WSNs) that integrates reinforcement learning (RL), supervised learning, and swarm intelligence techniques such as genetic algorithms (GA) and particle swarm optimization (PSO). Unlike conventional approaches that rely on static or standalone algorithms, the proposed framework employs a structured decision-making pipeline that dynamically adapts to real-time changes in network topology, traffic, and energy conditions. Each AI module plays a distinct role-RL handles local routing decisions, while GA and PSO are invoked for global optimization under resource constraints. Simulations conducted in MATLAB R2021b validate the framework's effectiveness, demonstrating improvements in packet delivery ratio, end-to-end latency, and energy efficiency when compared to traditional protocols. While this study is based on synthetic evaluations, it outlines the architectural groundwork for future real-world implementation and discusses deployment challenges such as scalability, resource usage, and security. The results highlight the potential of hybrid AI-based routing strategies to enhance the reliability, adaptability, and sustainability of WSNs in dynamic and resource-limited environments.
本文提出了一种基于模块化人工智能（AI）的无线传感器网络（WSN）路由框架，该框架集成了强化学习（RL）、监督学习以及遗传算法（GA）和粒子群优化（PSO）等群体智能技术。与依赖静态或独立算法的传统方法不同，所提出的框架采用了一种结构化决策管道，可动态适应网络拓扑、流量和能量状况的实时变化。每个人工智能模块都发挥着独特的作用——强化学习处理本地路由决策，而遗传算法和粒子群优化则在资源约束下用于全局优化。在MATLAB R2021b中进行的仿真验证了该框架的有效性，与传统协议相比，在数据包传输率、端到端延迟和能源效率方面均有改进。虽然本研究基于综合评估，但它概述了未来实际应用的架构基础，并讨论了诸如可扩展性、资源使用和安全性等部署挑战。结果突出了基于混合人工智能的路由策略在动态和资源受限环境中增强无线传感器网络可靠性、适应性和可持续性的潜力。
