Near-field ISAC for a RIS-assisted system

A novel reconfigurable intelligent surfaces (RIS) assisted near-field (NF) ISAC system is investigated, where the spherical wave propagation environment is utilized to elevate the radio sensing performance. Except for the conventional RIS, the sensor elements are embedded on the RIS surface to conduct the radios sensing functionality. By exploiting the symmetry property of the steering vector, a new expression of position error bound (PEB) is derived to unveil the impact of the sensor deployment. To balance the radio sensing performance and the sensor deployment cost, a cost function minimization problem is formulated to jointly optimize the number of sensor elements and the passive beamforming (BF). To solve this non-convex problem, a joint geometric programming element-wise (JGPE) algorithm is proposed. The successive convex approximation for geometric programming is invoked to optimize the number of sensor elements while the element-wise algorithm is adopted to optimize the passive BF. Numerical results demonstrated that the proposed algorithm reach the least PEB and cost function value among the benchmarks.