Efficient Beam Search For SatelliteCommunications On The Move Systems

Abstract Satellite Communication on the Move (SOTM) is essential for dynamic environments, such as military operations and disaster relief. Conventionalsystems with parabolic reflector antennas suffer from slow mechanical scanningand reduced tracking efficiency, especially for low Earth orbit (LEO) satellites,while phased array antennas offer faster, more precise tracking but remain underutilized. A key challenge in SOTM is achieving precise beam alignment betweenthe satellite and moving terminals, which directly impacts tracking accuracy.Several traditional existing beam alignment techniques are computationallyexpensive and impractical for real-time use. In this work, we propose a novel andefficient beam alignment algorithm that exploits the inherent low-rank structureof the beamspace energy matrix. By modeling the alignment process as amaximization problem with non-convex constraints, we leverage randomizedmatrix approximation to reconstruct a low-rank representation of the beamspace.This allows us to identify a subspace containing the optimal beam direction. Wethen use Particle Swarm Optimization (PSO) to refine the beam directionestimate. As illustrated in the numerical simulations, our approach improvesbeam alignment accuracy while reducing computational complexity comparing totraditional methods. This advancement has broad implications for enhancing theperformance of SOTM systems in dynamic environments.