An Energy‐Efficient Localization Framework for Mobile Wireless Sensor Networks Using Neural Attention–Guided Stochastic Fractal Search

ABSTRACT The use of wireless sensor networks (WSNs) for location‐based services is gaining importance in situations where conventional positioning systems are ineffective, such as GPS. Mobile and large‐scale deployments, in particular, pose a significant challenge in terms of maintaining high localization accuracy while maintaining energy efficiency. In this paper, an adaptive framework for energy‐efficient WSN on mobile devices is presented. A cluster‐based architecture is integrated with Particle Swarm Optimization (PSO) to optimize power allocation and minimize energy consumption. The system model incorporates realistic assumptions, including channel fading, received signal strength variations, and anchor node location uncertainties. The proposed approach consistently outperforms existing models such as FLA‐RTWOA, RDEANTN, and QLAMSR when measured on metrics such as localization accuracy, energy efficiency, and network lifetime. Simulation results show improved scalability, higher accuracy, better energy management, and extended network lifetime across varying node densities. The proposed framework therefore offers a robust and practical solution for energy‐aware localization in modern wireless sensor network applications.