A Bistatic Sensing System in Space-Air-Ground Integrated Networks

TL;DR

This work investigates bistatic sensing in space-air-ground integrated networks (SAGIN) by placing a multi-antenna LEO satellite as the transmitter and a ground base station as the radar receiver to sense a target aircraft. The authors formulate a non-convex fractional program to maximize the SINR of the target echo through joint design of transmit beamforming ${\mathbf{t}}$ and receive filtering ${\mathbf{u}}$, solved via alternating optimization with closed-form updates: a fractional-programming-based solution for ${\mathbf{t}}$ and a generalized Rayleigh quotient-based solution for ${\mathbf{u}}. They derive efficient, closed-form expressions for both subproblems and prove convergence, with simulation results showing that receive filtering dominates performance, especially at higher satellite altitudes, and that increased transmit power and more antennas further improve SINR. The findings provide practical SAGIN design insights on where to allocate resources (antenna count, power, altitude) to optimize bistatic sensing performance. Overall, the paper demonstrates that bistatic SAGIN sensing can outperform traditional transceiver setups under realistic channel models and constraints.

Abstract

Sensing is anticipated to have wider extensions in communication systems with the boom of non-terrestrial networks (NTNs) during the past years. In this paper, we study a bistatic sensing system by maximizing the signal-to-interference-plus-noise ration (SINR) from the target aircraft in the space-air-ground integrated network (SAGIN). We formulate a joint optimization problem for the transmit beamforming of low-earth orbit (LEO) satellite and the receive filtering of ground base station. To tackle this problem, we decompose the original problem into two sub-problems and use the alternating optimization to solve them iteratively. Using techniques of fractional programming and generalized Rayleigh quotient, the closed-form solution for each sub-problem is returned. Simulation results show that the proposed algorithm has good convergence performance.Moreover, the optimization of receive filtering dominates the optimality, especially when the satellite altitude becomes higher, which provides valuable network design insights.

Figures (4)

• Figure 1: An illustration of the space-air-ground sensing system.
• Figure 2: Convergence validation.
• Figure 3: Optimal SINR VS satellite altitude.
• Figure 4: Optimal SINR VS satellite antenna number.