A Computationally Efficient Beamforming Solution for Bistatic RF Sensing with RIS-Enhanced Target Illumination

Reconfigurable intelligent surface (RIS) is gaining increasing interest for communications and sensing in urban/indoor environments, where line-of-sight (LOS) propagation is often obstructed by buildings, walls, and other radio frequency (RF)-blocking objects. We consider a bistatic RF sensing system that employs RIS to provide enhanced target illumination and anti-blocking capabilities. The core problem is the joint design of the active transmit beamformer and the passive RIS beamformer, which is challenging due to nonconvex constant modulus constraints on the latter. Conventional approaches typically resort to convex relaxation or direct optimization on the constant-modulus manifold, which are computationally intensive because of their iterative nature. By exploiting the geometric structures inherent in the problem, we derive an analytic solution in closed form. In addition, we examine target detection with the RIS-assisted sensing system and quantitatively analyze the detection performance when the propagation paths in the system are subject to blocking. Finally, the RIS-assisted system is evaluated in various scenarios, including different RIS-sensor locations, channel estimation errors, and multipath interference.