.png){kind=icon}
.png){kind=icon}
Return
Semantic Communication Over MIMO Channels via Score-Based Reverse Mean Propagation
.png){kind=icon}
PRE
AI
DOI：10.1109/twc.2026.3703886.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This paper introduces a novel multiple-input multiple-output (MIMO) semantic communication system by leveraging the advanced score-based generative model. The proposed system features a deep neural network (DNN)-based encoder at the transmitter and a score-based decoder at the receiver, where the latter directly reconstructs source data from channel outputs without intermediate signal detection. We establish a unified Bayesian formulation for transceiver design rooted in the information-maximization principle. Based on this formulation, we propose a novel receiver framework which directly reconstructs the source data from raw channel outputs. We formulate a variational inference problem for receiver design and develop a principled score-based decoding algorithm, which generalizes reverse mean propagation (RMP) by incorporating the coding and channel constraints in likelihood calculations. With the proposed score-based receiver, we further discuss the channel-adaptive transmitter design with known channel state information (CSI) by combining channel-adaptive encoding with learned precoding. Experimental results on the FFHQ dataset demonstrate significant improvements in perceptual metrics over existing benchmarks, validating the effectiveness of the proposed approach in achieving high-fidelity image transmission.
Keywords:
Semantic communication
MIMO communication
score-based generative model
Journal
IF:
10.7
Papers:
1.3W
Citations:
5.3W