Near-Field Channel Estimation for Movable Antenna Systems: Angle-to-Distance Recovery via Movement Range Adjustment

Abstract

Movable antenna (MA) is an emerging technology that enables flexible adjustment of antenna positions to fully exploit continuous spatial variations of wireless channels. As the antenna movement region expands, the available spatial degrees of freedom (DoFs) increase, motivating the use of large movement regions. However, such large configurations inherently intensify near-field effects, rendering the conventional far-field assumption invalid. Recent studies have thus explored MA-enabled near-field communications, whose successful implementation critically relies on accurate channel state information (CSI). Nevertheless, most existing channel estimation techniques fail to reliably provide near-field CSI for MA systems, as they are primarily designed under the far-field assumption. A fundamental challenge lies in the strong spatial-domain coupling between angular and distance parameters, making their joint estimation a computationally prohibitive high-dimensional problem. To address this challenge, this paper presents an efficient near-field channel estimation framework for MA systems. Specifically, an MA movement-range adjustment strategy is introduced, where the MA is initially constrained to a small inner region so that far-field planar-wave behavior prevails, after which it traverses the full aperture to elicit the near-field spherical-wave response. Based on this staged movement strategy, an angle-to-distance orthogonal matching pursuit (A2D-OMP) algorithm is developed to sequentially recover angular and distance parameters. Furthermore, an angle–distance alternating gradient optimization (AD-AGO) algorithm is proposed to refine the initial estimates obtained by A2D-OMP. Numerical results demonstrate that the proposed framework significantly outperforms existing methods in accuracy while maintaining manageable computational complexity.