Adaptive Power Allocation and Link Selection in UAV-Assisted Hybrid FSO/RF Systems: Exploiting Weather-Dependent Performance Reversals

Hybrid UAV-assisted free-space optical (FSO)/radio-frequency (RF) links can deliver high throughput and robustness; however, they suffer from turbulence, rain, and pointing errors (PE).We present an adaptive power-sharing and link-selection strategy, and derive the closed-form bit error rate (BER) and outage expressions for exponentiated-Weibull optical turbulence and Nakagami-m RF fading with explicit pointing-error modeling.Monte Carlo simulations under clear air, light rain, and heavy rain were performed to validate the analysis.We reveal a power-/weather-dependent crossover: at low transmit power, the RF branch in light rain can outperform clear-air FSO, whereas at higher power, FSO dominates, and heavy rain suppresses this benefit.UAV relaying further reduces outages by up to five orders of magnitude in lightrain regimes, enhancing reliability.The contributions of this study are unified closed-form BER/outage analysis for exponentiated Weibull-Nakagami channels with PE; quantitative characterization of the power-/weather-dependent crossover with design implications; and an adaptive UAV-assisted framework enabling energy-aware power allocation and weather-aware link switching.