On the Performance of Vehicular Full-Duplex ISAC Systems With Cluster-Based Sensing Models

In-band Full-Duplex (FD) Multiple-Input Multiple-Output (MIMO) systems present a major opportunity for Integrated Sensing and Communication (ISAC) by enabling simultaneous signal transmission and reception. In this paper, we study the performance and implementation of FD-ISAC for vehicle-to-infrastructure (V2I) communications using a more precise sensing model. A common assumption in ISAC performance analysis has been that all targets can be represented by asingle pointcoordinate - named single ray sensing model (SRSM). This assumption simplifies performance analysis and algorithm implementation for ISAC, but it overlooks the shape and size of the sensing targets. Thus, we model each sensing target as acluster of rayswith certain angular characteristics associated with its geometry - named cluster ray sensing model (CRSM). Then, we derive the Cramer-Rao Bound (CRB) for angular direction and the variance of cluster-modeled targets. We also provide an (easy to optimize) lower bound of CRB for angular direction in terms of eigenvalues of the covariance matrix and its derivative, which approaches the original value in the low signal-to-noise (SNR) and low angular spread regime. Our analysis shows that a more accurate model (with large angular variance) improves the estimation of the direction of arrival of targets. Based on the CRB expressions, we design a partially connected hybrid beamforming (HBF) algorithm for millimeter wave (mmWave) OFDM FD-ISAC systems. The simulation results show that assuming CRSM together with a proposed HBF increases the data rate of the system by 25% as compared to assuming SRSM and a max signal to interference plus noise ratio (SINR) scheme.