Development of an Integrated Artificial Intelligence-Based Adaptive Rain Fade Mitigation Framework for High-Availability Ku-/Ka-Band Satellite Communication in Tropical Coastal Environments: A Case Study of Calabar, Nigeria

Abstract Rain fade remains one of the principal factors limiting the reliability and availability of satellite communication systems operating in tropical coastal regions characterized by intense precipitation. This study presents an integrated Artificial Intelligence (AI)-based adaptive rain fade mitigation model for enhancing the availability of high-capacity satellite communication links, using Calabar, Nigeria, as a case study. The proposed framework combines the ITU-R P.618 rain attenuation model, ITU-R P.838 specific attenuation model, adaptive uplink power control (AUPC), adaptive coding and modulation (ACM), and an AI-based attenuation prediction module to provide real-time mitigation of rain-induced signal degradation. A practical numerical design was developed for a 12 GHz Ku-band satellite link using a rain rate of 120 mm/h, an elevation angle of 45°, and a geostationary satellite distance of 38,000 km. The numerical analysis produced a free-space path loss of 205.63dB, a specific rain attenuation of 6.36 dB/km, an effective rain path length of 4.52 km, and a total rain attenuation of 28.75dB. Under severe rainfall conditions, the received signal power decreased to −  107.38dBm, falling below the receiver sensitivity threshold of −  95dBm, thereby requiring an adaptive compensation margin of 12.38dB. The AI prediction model estimated rain attenuation with a prediction error of only 0.35dB, achieving an RMSE of 0.58 dB, an MAE of 0.41 dB, and a coefficient of determination (R²) of 0.988, demonstrating excellent predictive accuracy. The adaptive controller dynamically increased the transmit power to 67.10 dBm and selected more robust modulation and coding schemes during severe fading conditions, thereby maintaining reliable communication. Graphical results illustrated the nonlinear increase of rain attenuation with rainfall intensity, the reduction of attenuation with increasing elevation angle, the decrease in received signal power during heavy rainfall, the convergence of the AI training process, the close agreement between measured and AI-predicted attenuation, the reduction in prediction error, the adaptive variation of transmit power and modulation schemes, and significant improvements in bit error rate, throughput, and overall system availability. Specifically, the proposed AI-based mitigation framework increased annual link availability from 99.60% to 99.96%, while maintaining stable communication under intense tropical rainfall. The integration of physics-based propagation modeling with artificial intelligence and adaptive communication techniques provide an effective, accurate, and computationally efficient solution for improving satellite communication reliability in tropical coastal environments.