GRU-based Channel Estimation with Hybrid FER-based Switching for UAV-to-Ground FSO Communication

Free space optical (FSO) communication has emerged as an attractive technology for high-data-rate wireless links, particularly for unmanned aerial vehicle (UAV)-to-ground communication scenarios, due to its advantages of unlicensed spectrum, narrow beam width, enhanced security, and minimal electromagnetic interference. However, despite these benefits, the performance of UAV-FSO links is severely affected by atmospheric turbulence, which introduces random fluctuations in the refractive index along the propagation path, leading to signal fading, intensity scintillation, beam wander, and beam spread. These effects degrade the received signal quality, resulting in increased bit error rates (BER) and reduced throughput. Traditionally, to ensure reliable data transmission under such varying channel conditions, Adaptive Coding and Modulation (ACM) techniques have been employed. ACM dynamically adjusts the modulation order and coding rate based on the current channel conditions to maximize throughput while ensuring an acceptable error performance. The effectiveness of ACM depends heavily on two fundamental components: (1) accurate channel state estimation and (2) robust switching decision mechanisms. In traditional UAV-to-ground FSO communication systems, these components are implemented using conventional estimation techniques and SNR-based switching standards.

Keywords: UAV, FSO communication, adaptive coding and modulation, GRU estimator, hybrid switching, LDPC, FER, turbulence mitigation.