Joint Localization and Channel Estimation Framework for FA-RIS Assisted ISAC

Artikel i vetenskaplig tidskrift, 2026

This paper proposes a fluid antenna (FA)-based channel reconstruction (FACR) framework for FA-reconfigurable intelligent surface (RIS) systems to improve channel estimation accuracy, inspired by the concept of integrated sensing and communication. Firstly, optimal FA placements for localization are determined based on the derived Cram´er-Rao lower bounds (CRLBs). By integrating the optimal FA placements and massive RIS with a proposed dual one-dimensional multiple signal classification (D1D-MUSIC) algorithm, the angle estimation performance closely approaches the CRLBs. Secondly, a complex expectation maximization-based refined estimation (CERE) algorithm is proposed to refine channel state information (CSI) and estimate channel statistical properties (CSPs). Thirdly, based on the estimated CSPs, a projection gradient-based pilot optimization (PGPO) algorithm is proposed to further improve channel estimation precision with optimized pilot symbols. Finally, simulation results demonstrate that D1D-MUSIC with optimized FA placements achieves close-to-bound performance. The CERE algorithm provides channel estimation performance nearly identical to that of an ideal optimal estimator, while achieving CSP estimation accuracy approaching the CRLBs. Furthermore, PGPO achieves further signal-to-noise ratio improvement for channel estimation with optimized pilot.