One-Step Generative Channel Estimation via Average Velocity Field
Zehua Jiang, Fenghao Zhu, Siming Jiang, Chongwen Huang, Zhaohui Yang, Richeng Jin, Zhaoyang Zhang, Merouane Debbah
TL;DR
The paper tackles the latency challenge of generative channel estimation in MIMO systems by introducing an average velocity field (AVF) framework that yields one-step channel estimation. By learning a direct mapping via the AVF, the method bypasses iterative denoising used in diffusion models, enabling fast inference with high accuracy. Experiments on 3GPP SCM data show AVF surpasses diffusion baselines in NMSE (up to 2.65 dB at high SNR) and reduces latency by about 90%, with robust performance across SNRs. The approach leverages angular-domain sparsity, a Jacobian-vector-product-based training, and supports multi-NFE for flexible refinement, marking a practical advancement for real-time wireless channel estimation.
Abstract
Generative models have shown immense potential for wireless communication by learning complex channel data distributions. However, the iterative denoising process associated with these models imposes a significant challenge in latency-sensitive wireless communication scenarios, particularly in channel estimation. To address this challenge, we propose a novel solution for one-step generative channel estimation. Our approach bypasses the time-consuming iterative steps of conventional models by directly learning the average velocity field. Through extensive simulations, we validate the effectiveness of our proposed method over existing state-of-the-art diffusion-based approach. Specifically, our scheme achieves a normalized mean squared error up to 2.65 dB lower than the diffusion method and reduces latency by around 90%, demonstrating the potential of our method to enhance channel estimation performance.