On the Performance Analysis of DL-Based Data Detection Algorithms in M-QAM Satellite Links

This work presents an analysis of deep learning (DL)-based data detection algorithms in the context of satellite communications, specifically considering challenging atmospheric environments (CAE). Since the satellite-to-earth communication systems migration to high-frequency bands, the influence of rain, fog, clouds, and vapors becomes more pronounced, imposing challenges for channel estimation. Thus, the adoption of DL-based data detection algorithms offers an alternative approach to improve the end-to-end performance. In our work, the performance of these algorithms is evaluated using the bit error rate (BER) metric in CAEs, considering Ku and K frequency bands. To the best of the author's knowledge, this study is the first to adopt and investigate the impact of atmospheric effects on DL-based data detection.