AI-Driven Efficient Downlink Communication for Limited-Transmit-Power CubeSats in the Ka-Band

CubeSats are miniature satellites that offer cost-effectiveness and versatility for space missions such as Earth observation, scientific experiments, and communication. However, their limited transmission power poses challenges for maintaining reliable communication, particularly in high-bandwidth Ka-band systems, where rain fading can significantly degrade the signal-to-noise ratio (SNR). This attenuation can reach up to 10 dB, reducing the quality of the communication link. Adaptive Modulation and Coding (AMC) schemes can dynamically adjust modulation and coding based on real-time SNR conditions, optimizing data rates despite fluctuating SNR. However, existing AMC algorithms, such as DVB-S2, need modification to meet the unique requirements of low-power CubeSat systems. Incorporating binary modulation schemes and simplifying the number of modulation and coding combinations can reduce design complexity while maintaining reliable links. This research aims to gather an extensive SNR dataset under various weather conditions and elevation angles for low-power CubeSats. We propose an AI-based algorithm that predicts SNR values over periods longer than the satellite signal propagation time, enabling proactive modulation and coding adjustments. The predicted SNR will allow the AMC system to optimize data transmission, reducing outages and improving communication efficiency. Our AI-driven approach for CubeSat AMC in the Ka-band offers a promising solution to overcome the challenges posed by high attenuation and fluctuating SNR. This work will significantly enhance the reliability of CubeSat communication systems, supporting more ambitious space missions and providing valuable insights into future satellite communications.