TaskCommMCR2

This repository is the official implementation of the paper:

• Multi-Device Task-Oriented Communication via Maximal Coding Rate Reduction [IEEE TWC] [arXiv]
• Authors: Chang Cai (The Chinese University of Hong Kong), Xiaojun Yuan (University of Electronic Science and Technology of China), and Ying-Jun Angela Zhang (The Chinese University of Hong Kong)

Brief Introduction

Existing Studies: Inconsistent Objectives for Learning and Communication

This Work: Synergistic Alignment of Learning and Communication Objectives

Usage

Feature Encoding

• Download images and put it under modelnet40_images_new_12x: Shaded Images (1.6GB). If the link does not work, you can download it from my Google Drive backup.

• Set environment: code is tested on python 3.7.13 and pytorch 1.12.1.

• Run the script main_train_phase1.py for the first-phase training of feature encoding. Then, run the script main_train_phase2.py for the second-phase training of feature encoding. Check mvcnn_pytorch for the details of the two training phases.

• Alternatively, download the pretrained checkpoints at Google Drive. The checkpoints can be used for feature extraction by running the script feature_load.py.

Precoding Optimization and Performance Evaluation

The code is located at the folder precoding_opt_matlab. Run the script main_precoding_opt.m to compare the performance of the proposed MCR² precoder and the LMMSE precoder.

Citation

If you find our work interesting, please consider citing

@ARTICLE{task_comm_mcr2,
  author={Cai, Chang and Yuan, Xiaojun and Zhang, Ying-Jun Angela},
  journal={IEEE Transactions on Wireless Communications}, 
  title={Multi-Device Task-Oriented Communication via Maximal Coding Rate Reduction}, 
  year={2024},
  volume={23},
  number={12},
  pages={18096-18110}
  }

Our follow-up work provides an information-theoretic interpretation of the learning-communication separation, as well as an end-to-end learning framework:

@ARTICLE{info_theoretic_e2e,
  author={Cai, Chang and Yuan, Xiaojun and Zhang, Ying-Jun Angela},
  journal={IEEE Journal on Selected Areas in Communications}, 
  title={End-to-End Learning for Task-Oriented Semantic Communications Over {MIMO} Channels: An Information-Theoretic Framework}, 
  year={2025},
  volume={},
  number={},
  pages={1-16}
  }