Edge Computing Resource Allocation for UAV Swarm via Improved Grey Wolf Optimization

Major earthquakes frequently damage critical infrastructure like base stations, disrupting communication networks and hindering search and rescue (SAR) operations. Unmanned Aerial Vehicle (UAV) swarms with edge computing offer an effective solution for rapid temporary network deployment in such scenarios. This paper proposes an improved Grey Wolf Optimizer (GWO) for optimal resource allocation in post-disaster UAV swarms by damaged base stations. Integrated communication, computation, and energy consumption models are developed, incorporating task partitioning and processing queues. A joint optimization problem minimizes system latency and energy consumption by co-optimizing task offloading ratios, transmission power, and bandwidth allocation. The improved GWO incorporates logistic chaotic initialization, prior-knowledge pruning and adaptive search strategies. Simulation results demonstrate that the proposed algorithm achieves superior energy efficiency compared to benchmark algorithms.