Performance Analysis of STAR-IRS-Aided MISO-ISAC Systems With Multiple Targets: A Rate-Splitting Approach
Soumen Mondal, Keshav Singh, Cunhua Pan, Chih–Peng Li
Abstract
The paper evaluates the ergodic sum capacity, outage performance for communication users, and the detection probability, beampattern gain for sensing targets in a simultaneous transmitting and reflecting intelligent reflecting surfaces (STAR-IRS) aided integrated sensing and communication (ISAC) system. The rate splitting multiple access (RSMA) technique and maximal ratio transmit beamforming at the multi-antenna base station have been explored. Closed-form expressions for the ergodic sum capacity and outage probability of the STAR-IRS-aided RSMA ISAC system are derived through moment methods. The derived expressions are validated through Monte Carlo simulations. Additionally, to provide deeper insights into the diversity orders of the RSMA ISAC system, we conduct an asymptotic outage probability analysis in the high signal-to-noise ratio regime. The effect of the number of base station antennas and STAR-IRS elements on outage performance has been demonstrated, along with an explanation of the underlying reasons through diversity gain. Furthermore, it shows that the implementation of STAR-IRS significantly boosts the system’s ergodic sum capacity compared to traditional reflecting-only IRS. Additionally, the RSMA technique delivers more substantial performance improvements than the non-orthogonal multiple access (NOMA) in high transmit SNR conditions while demonstrating comparable performance in low transmit SNR scenarios. A comparison between energy splitting and mode switching STAR-IRS has been conducted under both ideal and random phase shift conditions. A trade-off analysis between communication and sensing rates is presented. Additionally, the accuracy of target sensing is evaluated by measuring the mean square error (MSE) in beampattern gain matching. The impact of quantization levels for phase shift of STAR-IRS on outage probability has also been addressed. Finally, the effects of power allocation for sensing on detection probability and beam pattern gain are also presented.