Randomized trial explores AI-enhanced adaptive fade mitigation in satellite systems, suggesting improved quality of service.
Key Points
The research aims to enhance adaptive fade mitigation techniques for satellite communications by leveraging AI algorithms to predict channel conditions.
Proposes the use of LSTM deep learning models to predict rain attenuation using real data from Chilbolton and Chilton Observatories.
Integrates time diversity, site diversity, selection combining, and maximal ratio combining into an AFMT strategy.
Trained and validated the model using MATLAB R2020b, achieving simulation results that indicate significant performance improvements.
LSTM model achieved 90% accuracy in predicting rain attenuation for 0.01% time probability.
Hybrid AFMT strategy delivered at least 21 dB gain compared to other configurations.
Highlights potential of AI-driven solutions for enhancing signal reliability and quality in satellite communications.