COVERAGE OPTIMIZATION IN UAV-ASSISTED DRL BASED WIRELESS SENSOR NETWORK FOR DISASTER MANAGEMENT

Abstract

Unmanned aerial vehicle integrated with wireless sensor network is efficient in managing the sensor node connectivity as well as data transmission during the existence of disaster, the maximum area coverage and data transmission are ensured especially when nodes are grouped into clusters. Wireless Sensor Networks (WSNs) play a vital role in large-scale environmental monitoring, surveillance, and disaster management. However, the inherent energy constraints and limited communication range of sensor nodes pose significant challenges in ensuring efficient coverage and data collection. These limitations are overcome by using this novel proposal, this paper proposes a novel framework integrating Unmanned Aerial Vehicles (UAVs) with Deep Reinforcement Learning (DRL) for coverage optimization in clustered WSNs, each managed by a Cluster Head (CH), which aggregates data and communicates with UAVs. A DRL agent is trained to dynamically plan UAV trajectories and scheduling policies that maximize coverage and data collection efficiency while minimizing energy consumption and communication delays. The state space captures real-time UAV positions, CH locations, and node energy levels, while the reward function encourages full cluster coverage, reduced redundant visits, and balanced load distribution. Simulation results show that the proposed DRL-based UAV-assisted strategy significantly improves sensing coverage, reduces data latency, and extends network lifetime compared to traditional approaches.