Spatially Non-Stationary XL-MIMO Channel Estimation: A Three-Layer Generalized Approximate Message Passing Method

Anzheng Tang,Jun-Bo Wang,Yijin Pan,Wence Zhang,Yijian Chen,Hongkang Yu,Rodrigo C. de Lamare
2024-09-20
Abstract:In this paper, channel estimation problem for extremely large-scale multi-input multi-output (XL-MIMO) systems is investigated with the considerations of the spherical wavefront effect and the spatially non-stationary (SnS) property. Due to the diversities of SnS characteristics among different propagation paths, the concurrent channel estimation of multiple paths becomes intractable. To address this challenge, we propose a two-phase channel estimation scheme. In the first phase, the angles of departure (AoDs) on the user side are estimated, and a carefully designed pilot transmission scheme enables the decomposition of the received signal from different paths. In the second phase, the subchannel estimation corresponding to different paths is formulated as a three-layer Bayesian inference problem. Specifically, the first layer captures block sparsity in the angular domain, the second layer promotes SnS property in the antenna domain, and the third layer decouples the subchannels from the observed signals. To efficiently facilitate Bayesian inference, we propose a novel three-layer generalized approximate message passing (TL-GAMP) algorithm based on structured variational massage passing and belief propagation rules. Simulation results validate the convergence and effectiveness of the proposed algorithm, showcasing its robustness to different channel scenarios.
Information Theory,Signal Processing
What problem does this paper attempt to address?
### Problems Addressed by the Paper The paper primarily aims to address the issue of channel estimation in extremely large-scale multiple-input multiple-output (XL-MIMO) systems, particularly considering the spherical wavefront effect and spatial non-stationary (SnS) characteristics. #### Main Challenges and Background - **XL-MIMO Systems**: With the development of wireless communication technology and the increasing demand for higher data rates, the number of antennas and array aperture significantly exceed those of existing large-scale MIMO systems. This leads to near-field transmission phenomena, necessitating a reevaluation of various channel characteristics. - **Spherical Wavefront Effect and SnS Characteristics**: Due to the large aperture antenna arrays introducing non-negligible Rayleigh distances, the plane wave assumption is no longer applicable. Additionally, SnS characteristics cause different propagation paths to exhibit varying behaviors across the antenna array, increasing the complexity of multipath channel estimation. #### Limitations of Existing Methods - Most existing methods are only applicable to spatially stationary channel models and fail to fully exploit the inherent structural characteristics of XL-MIMO channels. - Existing SnS channel estimation methods, such as subarray estimation schemes and scatterer estimation schemes, although providing preliminary solutions, do not fully explore the potential characteristics in the antenna or angle domain. #### Proposed Methods - **Two-Stage Channel Estimation Scheme**: First, the angles of departure (AoD) on the user side are estimated using super-resolution methods, then these estimates are used to decompose the overall channel estimation problem into multiple sub-channel estimation tasks. - **Three-Layer Bayesian Inference Framework**: To accurately address the second stage's VR and channel coefficient cascaded estimation problem, a novel three-layer Bayesian inference framework is proposed, capturing characteristics in the angle domain and antenna domain at different layers. - **TL-GAMP Algorithm**: A three-layer generalized approximate message passing (TL-GAMP) algorithm is developed to efficiently perform Bayesian inference, demonstrating superior performance and robustness in various channel scenarios. ### Summary The paper aims to solve the complex channel estimation problem in XL-MIMO systems, particularly considering the spherical wavefront effect and spatial non-stationary characteristics. By proposing a two-stage estimation scheme and a three-layer Bayesian inference framework, it effectively addresses the challenges in multipath channel estimation and demonstrates excellent performance in different environments.