Enhanced Polar-Domain Channel Estimation for Near-Field XL-MIMO in Low-SNR Scenarios
Han Wang, Tianyu Yan, Nanrun Zhou, Xingwang Li, Fangqing Wen, Wencai Du
Abstract
This paper presents an enhanced approach for polar-domain channel estimation in near-field extremely large scale multiple input multiple output (XL-MIMO) systems, focusing on improving performance under low signal-to-noise ratio (SNR) conditions. The proposed method combines a lazy residual update strategy with adaptive weight adjustments, optimizing estimation accuracy in challenging environments. Extensive simulations demonstrate that the proposed algorithm outperforms traditional compressed sensing techniques in both angle-domain and polar-domain settings, as well as the bilinear pattern detection (BPD) based algorithm. Compared to the multi-candidate BPD (MBPD) method, the proposed approach exhibits superior mean squared error (MSE) performance, particularly in low-SNR scenarios. Both methods employ adaptive weighting, but the lazy residual update reduces the frequency of full updates, enhancing computational efficiency without sacrificing accuracy. While not consistently better than MBPD in all cases, the proposed algorithm proves particularly effective under low-SNR conditions. These findings confirm its robustness and efficiency for near-field XL-MIMO systems.