Real-Time UAV Collaborative Beam Reforming for Coexistent Satellite-Terrestrial Communications

This paper presents a novel collaborative UAV beam reforming approach to dynamically form and adapt a space-selective beam path to coexist satellite and terrestrial communications and overcome the challenges of interference from the real-time satellite scheduling and other active/passive users present in the neighboring networks. The collaboration within a cluster of randomly distributed UAVs emulates the antenna arrays for dynamic beamforming and reforming, reducing interference in impacted zones with passive and active receivers of satellite communications. Different from conventional fixed-length mmWave beamforming, the distributed UAV beam reforming allows increased space among antennas, enables the narrowing of bandwidth for arrays, and improves the network's ability to focus energy in a specific direction. The hovering of the UAVs makes it challenging to accurately control the beam side-lobes and leads to distortion in real-time, which is crucial in avoiding interference with satellite-terrestrial communication and maximizing the dynamic spatial co-existence. Meanwhile, increased spacing within the UAVs results in higher sidelobes resulting in a reduced received power and higher latency at the receiver with increased interference to the active/passive users of the neighboring networks. Thus, we thoroughly assessed the impacts to system performance caused by the hovering and proposed a Q-Learning-based strategy that effectively reduces the hovering impacts, and significantly improves the beam communication quality. Considering the receiver's sensitivity to the angle of arrival (AoA), the proposed Q-learning algorithm based on the AoA information increases the hovering tolerance of a UAV to maintain a reliable performance or required target quality-of-service. Increased directivity distinguishes signals arriving from different directions at the receiver, thereby improving the receiver's ability to suppress interference and enhance the desired signal. Thus, the receiver and transmitter hovering stability are fed back instantly to the beam-forming and reforming engine to achieve the balance of the angle of arrival sensitivity and selectivity in real-time. Generated beam instability, impacted by the number of antennas and the spacing among antennas, is thoroughly assessed with theoretical performance and randomness in practice. Combining the theoretical results and more complicated environments in practice, the presented system enables aerial access points served by collaborative UAVs to maximize the coexistence of satellite and terrestrial communication. It addresses the challenges of interference and spectrum utilization in Space-Air-Ground Integrated networks (SAGIN).