Joint Channel Sounding and Source-Channel Coding for MIMO-OFDM Systems: Deep Unified Encoding and Parallel Flow-Matching Decoding

TL;DR

This paper introduces a deep unified encoder and a parallel flow-matching decoder for MIMO-OFDM systems, significantly improving joint channel and source estimation without explicit pilot-data separation.

Contribution

It presents a novel deep encoding and decoding framework that jointly estimates channel and source, outperforming existing methods in efficiency and accuracy.

Findings

Outperforms state-of-the-art in channel estimation accuracy

Achieves superior source reconstruction quality

Provides a new benchmark using Bayesian Cramér-Rao bound

Abstract

In this work, we propose a deep unified (DU) encoder that embeds source information in a codeword that contains sufficient redundancy to handle both channel and source uncertainties, without enforcing an explicit pilot-data separation. At the receiver, we design a parallel flow-matching (PFM) decoder that leverages flow-based generative priors to jointly estimate the channel and the source, yielding much more efficient inference than the existing diffusion-based approaches. To benchmark performance limits, we derive the Bayesian Cram\'er-Rao bound (BCRB) for the joint channel and source estimation problem. Extensive simulations over block-fading MIMO-OFDM channels demonstrate that the proposed DU-PFM approach drastically outperforms the state-of-the-art methods in both channel estimation accuracy and source reconstruction quality.