Variational Source-Channel Coding for Semantic Communication

Yulong Feng,Jing Xu,Liujun Hu,Guanghui Yu,Xiangyang Duan

2024-10-17

Abstract:Semantic communication technology emerges as a pivotal bridge connecting AI with classical communication. The current semantic communication systems are generally modeled as an Auto-Encoder (AE). AE lacks a deep integration of AI principles with communication strategies due to its inability to effectively capture channel dynamics. This gap makes it difficult to justify the need for joint source-channel coding (JSCC) and to explain why performance improves. This paper begins by exploring lossless and lossy communication, highlighting that the inclusion of data distortion distinguishes semantic communication from classical communication. It breaks the conditions for the separation theorem to hold and explains why the amount of data transferred by semantic communication is less. Therefore, employing JSCC becomes imperative for achieving optimal semantic communication. Moreover, a Variational Source-Channel Coding (VSCC) method is proposed for constructing semantic communication systems based on data distortion theory, integrating variational inference and channel characteristics. Using a deep learning network, we develop a semantic communication system employing the VSCC method and demonstrate its capability for semantic transmission. We also establish semantic communication systems of equivalent complexity employing the AE method and the VAE method. Experimental results reveal that the VSCC model offers superior interpretability compared to AE model, as it clearly captures the semantic features of the transmitted data, represented as the variance of latent variables in our experiments. In addition, VSCC model exhibits superior semantic transmission capabilities compared to VAE model. At the same level of data distortion evaluated by PSNR, VSCC model exhibits stronger human interpretability, which can be partially assessed by SSIM.

Signal Processing,Information Theory,Machine Learning

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: how to achieve efficient semantic communication through Joint Source - Channel Coding (JSCC), and explain why the existing Auto - Encoder (AE) methods are insufficient in handling semantic communication. Specifically, the paper focuses on the following points:

1. **Differences between semantic communication and classical communication**:
- Classical communication theories (such as Shannon's information theory) mainly focus on lossless transmission, that is, minimizing data distortion as much as possible. However, semantic communication allows a certain degree of data distortion in order to improve communication efficiency. This is because semantic information does not need to completely rely on all the details of the original data.
- Semantic communication introduces data distortion, which makes it different from classical communication. According to the rate - distortion theory, semantic communication can reduce the amount of transmitted data by introducing appropriate distortion, thus improving efficiency.

2. **Importance of Joint Source - Channel Coding (JSCC)**:
- Although the existing auto - encoder methods can be trained end - to - end, they perform poorly in dealing with channel characteristics and cannot clearly explain the role of the channel in joint coding.
- The paper proposes a Variational Source - Channel Coding (VSCC) method, aiming to integrate channel characteristics into the coding process through variational inference, so as to better handle the transmission of semantic information.

3. **Design of semantic communication systems**:
- The paper explores how to design a semantic communication system based on a deep - learning network, using the VSCC method to capture and transmit semantic features.
- The experimental results show that the VSCC model is superior to the traditional AE model and the Variational Auto - Encoder (VAE) model in terms of interpretability and semantic transmission ability.

4. **Theoretical analysis and experimental verification**:
- Through the research on semantic information theory, the paper proposes a new framework for semantic communication and verifies the effectiveness of the VSCC method through experiments.
- The experimental results show that, at the same distortion level, the VSCC model has stronger human interpretability and higher semantic transmission performance.

In summary, this paper aims to solve the problems existing in the current semantic communication systems, especially the deficiencies in dealing with channel characteristics and semantic information transmission, by introducing the Variational Source - Channel Coding method.