Hardware-Aware Reflection Element Selection and Parallel ISI Cancellation for RIS-FTN Systems

Abstract

High spectral efficiency and wide coverage enhancements for future wireless networks are offered by reconfigurable intelligent surfaces (RIS) and faster-than-Nyquist (FTN) signaling. However, severe challenges are faced in practical deployments from RIS hardware impairments and aggressive FTN-induced inter-symbol interference (ISI). A robust hardware-aware RISassisted FTN communication system is proposed in this paper. First, to mitigate hardware defects and reduce channel acquisition overhead, a reflection element selection (RES) strategy utilizing a maximum cascaded magnitude selection (MCMS) criterion is introduced. Second, to efficiently suppress severe ISI and colored noise, a parallel soft iterative interference cancellation (Soft-ISIC) detector is developed. Ergodic channel capacity and average bit error rate bounds are theoretically derived. Analytical derivations are validated through extensive simulations, demonstrating that near-optimal full-array performance with selective active elements is achieved by the MCMS-RES, while the performance bottleneck of traditional linear equalizers under severe hardware damage is successfully mitigated by the Soft-ISIC detector.