Successive Bayesian Reconstructor for Channel Estimation in Fluid Antenna Systems

TL;DR

This work introduces the Successive Bayesian Reconstructor (S-BAR) for channel estimation in fluid antenna systems (FASs), addressing model mismatch and pilot overhead by treating the FAS channel as a stochastic process and performing kernel-based Bayesian regression. S-BAR designs a switch matrix through greedy, mutual-information driven port selection and then reconstructs the channel as the posterior mean, with an offline-online two-stage implementation that yields linear online complexity in the number of ports. Theoretical analysis provides a minimum mean-squared error bound dependent on the eigenvalues of the true covariance, and simulations across QuaDRiGa and SSC channels show that S-BAR surpasses model-based estimators, achieving higher accuracy with fewer pilots and remains robust to kernel misspecification. The approach highlights the practical potential of prior-informed, non-parametric channel estimation for FASs, including resilience to EM coupling and significant pilot savings in realistic scenarios.

Abstract

Fluid antenna systems (FASs) can reconfigure their antenna locations freely within a spatially continuous space. To keep favorable antenna positions, the channel state information (CSI) acquisition for FASs is essential. While some techniques have been proposed, most existing FAS channel estimators require several channel assumptions, such as slow variation and angular-domain sparsity. When these assumptions are not reasonable, the model mismatch may lead to unpredictable performance losses. In this paper, we propose the successive Bayesian reconstructor (S-BAR) as a general solution to estimate FAS channels. Unlike model-based estimators, the proposed S-BAR is prior-aided, which builds the experiential kernel for CSI acquisition. Inspired by Bayesian regression, the key idea of S-BAR is to model the FAS channels as a stochastic process, whose uncertainty can be successively eliminated by kernel-based sampling and regression. In this way, the predictive mean of the regressed stochastic process can be viewed as a Bayesian channel estimator. Simulation results verify that, in both model-mismatched and model-matched cases, the proposed S-BAR can achieve higher estimation accuracy than the existing schemes.

Figures (8)

• Figure 1: An illustration of channel estimation for an fas, where one $N$-port bs equipped with $M$ fluid antennas receives pilots from a user in the uplink.
• Figure 2: The frame structure for fas channel reconstruction.
• Figure 3: An illustration of employing S-BAR scheme to estimate fas channel $\bf h$. (a)-(d) illustrate the real part of $\bf h$ versus the index of ports. (e)-(h) illustrate the imaginary part of $\bf h$ versus the index of ports.
• Figure 4: The nmse as a function of the receiver snr under the assumption of QuaDRiGa channel model.
• Figure 5: The nmse as a function of the receiver snr under the assumption of SSC channel model.