IRS-UAV assisted military communications in mountainous NLOS environments: A hybrid metaheuristic anti-jamming strategy

Protecting military communications in mountainous Non-Line-Of-Sight (NLOS) areas is difficult due to active jamming. Using an Intelligent Reflecting Surface (IRS) on an Unmanned Aerial Vehicle (UAV) is a potential solution by uniquely combining the UAV’s 3D mobility with the passive, energy-efficient signal reconfiguration capabilities of the IRS to actively evade and suppress interference. However, jointly optimizing the UAV flight path and IRS settings is complex. This study proposes a framework to solve this challenge using a hybrid algorithm called Grey Wolf-Artificial Rabbits Optimization (GWARO). Featuring a median-based dynamic thresholding mechanism to autonomously balance exploration and exploitation, the algorithm coordinates a dual-layer defense: physical terrain masking and electronic nulling to suppress interference. System simulations utilized 3D Shuttle Radar Topography Mission (SRTM) data and Rician fading models. Under a 200 W jamming threat with a 512-element IRS, GWARO achieves an average Signal-to-Jammer Ratio (SJR) of 8.34 dB and a 95.0% success rate. Although GWARO is approximately 14% slower than standard Particle Swarm Optimization (PSO), a significant trade-off is struck through its improved signal stability and enhanced jamming suppression. Furthermore, GWARO exhibits a superior computational profile against its parent algorithms, providing maximum speedups of 12.50% over Artificial Rabbits Optimization (ARO) and 10.97% over Grey Wolf Optimizer (GWO). Wilcoxon tests confirmed these performance improvements as significant ( p < 0 . 05 ). Findings confirm the proposed framework as a reliable solution for military networks, wherein communication link quality is of greater interest than computational cost.