Predictive Beamforming for Joint Radar and Communication Transmission in LEO Satellite Aeronautical Systems

Abstract

LEO satellite communication is essential to provide aviation terminal (AT) services. However, due to the relative motion between LEO satellites and ATs, ATs are required to frequently report their current positions to the satellite to guarantee continuous service, which may lead to significant signaling overhead. To overcome this issue, we propose a predictive beamforming approach based on location sensing, whose core idea is to leverage a unified beampattern to enable sensing and data transmission simultaneously. Specifically, Cramér-Rao lower bound (CRLB) is adopted to characterize position tracking accuracy under the extended Kalman filtering framework. Then, the concerned problem is formulated to optimize spectral efficiency under transmission power and tracking accuracy constraints. To deal with the problem, we first prove the monotonic relationship between tracking accuracy and the perceived signal-to-interference-plus-noise ratio at the satellite to reformulate the problem, and then it is solved optimally by combining branch-and-reduce-and-bound approach with second-order cone programming. Towards practical application, the primal problem is equivalently reformulated with quadratic transform, and then semidefinite relaxation is adopted for low-complexity beamformer design. Simulation results verify the superior tracking and communication performance of our low-complexity design, and it is shown that it achieves comparable transmission rate and tracking accuracy relative to optimum. Moreover, the average spectral efficiency and tracking accuracy can be improved by up to 60% and 30% relative to other baselines.