Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Multi-user Wireless Image Semantic Transmission over MIMO Multiple Access Channels

Bingyan Xie, Yongpeng Wu, Feng Shu, Jiangzhou Wang, Wenjun Zhang

TL;DR

The paper tackles uplink semantic image transmission over a two-user MIMO-MAC by introducing MU-LCFSC, a CSI-aware framework that fuses channel state information into both semantic encoders and decoders and uses a cooperative SIC-based decoder with a CVAE-inspired mask-ratio generator. Its core contributions include a cooperative semantic decoder with a learnable mask ratio, and a CVAE-based mechanism (CMRG) to adapt attention masks for inter-user interference mitigation. Empirical results on the UDIS-D dataset show MU-LCFSC achieving about $3$ dB PSNR gains over DeepJSCC-NOMA under comparable conditions, with robust performance across varying SNRs and CBRs and improved visual quality. The approach promises enhanced robustness and spectral efficiency for multi-user semantic communications in 6G uplink scenarios, by jointly optimizing semantic encoding/decoding and interference cancellation under MIMO fading.

Abstract

This paper focuses on a typical uplink transmission scenario over multiple-input multiple-output multiple access channel (MIMO-MAC) and thus propose a multi-user learnable CSI fusion semantic communication (MU-LCFSC) framework. It incorporates CSI as the side information into both the semantic encoders and decoders to generate a proper feature mask map in order to produce a more robust attention weight distribution. Especially for the decoding end, a cooperative successive interference cancellation procedure is conducted along with a cooperative mask ratio generator, which flexibly controls the mask elements of feature mask maps. Numerical results verify the superiority of proposed MU-LCFSC compared to DeepJSCC-NOMA over 3 dB in terms of PSNR.

Multi-user Wireless Image Semantic Transmission over MIMO Multiple Access Channels

TL;DR

The paper tackles uplink semantic image transmission over a two-user MIMO-MAC by introducing MU-LCFSC, a CSI-aware framework that fuses channel state information into both semantic encoders and decoders and uses a cooperative SIC-based decoder with a CVAE-inspired mask-ratio generator. Its core contributions include a cooperative semantic decoder with a learnable mask ratio, and a CVAE-based mechanism (CMRG) to adapt attention masks for inter-user interference mitigation. Empirical results on the UDIS-D dataset show MU-LCFSC achieving about dB PSNR gains over DeepJSCC-NOMA under comparable conditions, with robust performance across varying SNRs and CBRs and improved visual quality. The approach promises enhanced robustness and spectral efficiency for multi-user semantic communications in 6G uplink scenarios, by jointly optimizing semantic encoding/decoding and interference cancellation under MIMO fading.

Abstract

This paper focuses on a typical uplink transmission scenario over multiple-input multiple-output multiple access channel (MIMO-MAC) and thus propose a multi-user learnable CSI fusion semantic communication (MU-LCFSC) framework. It incorporates CSI as the side information into both the semantic encoders and decoders to generate a proper feature mask map in order to produce a more robust attention weight distribution. Especially for the decoding end, a cooperative successive interference cancellation procedure is conducted along with a cooperative mask ratio generator, which flexibly controls the mask elements of feature mask maps. Numerical results verify the superiority of proposed MU-LCFSC compared to DeepJSCC-NOMA over 3 dB in terms of PSNR.

Paper Structure

This paper contains 17 sections, 13 equations, 5 figures, 1 table.

Table of Contents

  1. Introduction
  2. System Model and Proposed Framework
  3. System Model
  4. Proposed Framework of MU-LCFSC
  5. Cooperative Decoding Designs for the Multi-user Detection
  6. Implementation Details
  7. Numerical Results
  8. Experimental Setups
  9. Datasets
  10. Model Deployment Details
  11. Comparison Benchmarks
  12. Evaluation Metrics
  13. Results Analysis
  14. SNR Performances
  15. CBR Performances
  16. ...and 2 more sections

Figures (5)

  • Figure 1: MU-LCFSC framework with two-user pair over MIMO-MAC channels. Two users independently encode their images and decode them successively. The strong user decodes the codewords first and then the weak user utilizes the reconstructed results of strong user to recover the image.
  • Figure 2: The structure of cooperative mask ratio generator. Each mask ratio is learnt jointly through the chosen mask ratio range and value.
  • Figure 3: Quality of the reconstructed images versus the SNRs in MIMO fading channels ($R$ = 0.06).
  • Figure 4: Quality of the reconstructed images versus the CBRs in MIMO fading channels (SNR = 6 dB).
  • Figure 5: Visualized results for the MU-LCFSC and other benchmarks. (SNR = 12 dB, CBR = 0.10)