Large-scale Path Planning and Time Window Allocation in UAV-assisted Wireless Sensor Networks with Variational Multi-dimensional Optimization

Wireless Sensor Networks (WSNs) have been widely deployed, and it is crucial to enhance their network longevity and transmission efficiency. This work comprehensively considers communication interference, transmission rate constraints, and motion constraints of Unmanned Aerial Vehicles (UAVs). We propose an energy-efficient UAV-assisted WSN framework with big data transmission. To minimize the weighted sum of UAV's and WSNs' energy consumption, we jointly optimize sensor clustering, path planning, and time window allocation by an improved intelligent optimization algorithm assisted by deep learning named Variational Multidimensional Optimization (VMO) with co-evolved Multiple Subpopulations, noted as VMOMS for short. The effectiveness of the proposed VMOMS algorithm for solving high-dimensional problems is demonstrated through numerical analysis and simulation results. These findings highlight the efficiency and practicality of the designed UAV-assisted hierarchical architecture of WSNs, thereby showcasing its potential to enable reliable data transmission from remote WSNs to a centralized cloud server.