Position-Aware Hybrid Beamforming for ISAC: Leveraging RIS and Stacked Intelligent Metasurfaces

Although the stacked intelligent metasurface (SIM) has shown promise in enhancing integrated sensing and communication (ISAC) systems, its performance is fundamentally limited by blockages in practical scenarios. A natural solution is to introduce a reconfigurable intelligent surface (RIS) to alleviate this blockage problem. However, deploying the RIS brings new challenges, such as the “double fading” effect, which can severely degrade the end-to-end signal strength. To overcome these challenges, this paper investigates a novel ISAC framework that leverages SIM in conjunction with RIS to enable position-aware hybrid beamforming. We first derive the Cramér-Rao bound (CRB) for estimating the two-dimensional direction-of-arrival of the sensing target. Based on the proposed framework, the phase shifts of SIM and RIS, as well as the deployment location of the RIS, are jointly optimized to maximize the sum rate of users subject to a CRB constraint. However, this CRB constraint introduces a highly complex non-convexity with respect to the optimization variables, making direct optimization intractable. We therefore derive an intuitive relationship between the CRB and beampattern gain to enable tractable optimization. To tackle the resulting non-convex optimization problem, two efficient algorithms are developed: a high-performance algorithm termed SGE (based on successive convex approximation and gradient ascent with element-wise coordinate ascent), and a low-complexity method based on double gradient ascent with successive convex approximation (DGAS). Extensive simulation results demonstrate that introducing the RIS delivers a 1.5 times higher sum rate subject to the sensing constraint in challenging propagation conditions. Furthermore, RIS location optimization confers an additional gain relative to its fixed position counterpart.