Wireless Sensing with Movable Intelligent Surface

Le résumé fourni par la source

Future wireless networks are envisioned to deliver not only gigabit communications but also ubiquitous sensing. Reconfigurable intelligent surfaces (RISs) have emerged to reshape radio propagation, recently showing considerable promise for wireless sensing. Still, their per-element electronic tuning incurs prohibitive hardware cost and power consumption. Motivated by the concept of fluid antenna systems (FAS), this paper introduces a low-cost movable intelligent surface (MIS) for wireless sensing, which replaces element-wise electronic phase tuning with panel-wise mechanical reconfiguration. The MIS stacks two pre-phased metasurface (MS) layers: a large fixed layer and a smaller movable layer, whose differential position shifts synthesize distinct composite phase patterns, enabling multiple beam patterns for multi-target detection. We characterize a MIS-enabled multi-hop echo signal model with multi-target interference and then formulate a worst-case sensing signal-to-interference-plus-noise ratio (SINR) maximization problem that jointly designs MIS phase shifts and schedules MS2's position. A Riemannian Augmented Lagrangian Method (RALM)-based algorithm is developed to solve the formulated mixed-integer non-convex problem. We further extend the MIS sensing scheme to a peak-to-side-lobe-ratio (PSLR)-oriented design with the same algorithm framework. We also derive a heuristic MIS beam-steering design with a closed-form phase distribution and position scheduling. Simulations validate MIS's beam pattern reconfiguration capability, show that the RALM-based scheme significantly outperforms the closed-form scheme in improving sensing SINR, and uncover a gain-diversity trade-off in beam patterns that informs the optimal choice of MIS configuration.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.