IMM-UKF Model Based Adaptive Beam Tracking in XL-MIMO Systems

In the realm of 6 G wireless communications, Extremely Large-Scale Multiple-Input-Multiple Output (XLMIMO) emerges as a pivotal technology, enhancing directional signal transmission by employing hundreds to thousands of antennas. The integration of near-field communication and extremely narrow pencil beams in XL-MIMO systems poses significant challenges for Channel State Information (CSI) acquisition. To address these challenges, this paper introduces an Interacting Multiple Model with Gated Recurrent Units (IMM-GRU) model, which integrates multiple kinematic models to accurately capture the complex motion trajectories of users in near-field communication environments and efficiently assigns the optimal beam to users. This approach effectively addresses the limitations of existing beam tracking schemes that rely on uniform motion models and their inapplicability to XL-MIMO near-field communication systems. Simulation experiments demonstrate the robust tracking performance of the proposed IMM-GRU model, which is crucial for maintaining communication performance in dynamic environments.