Deep Learning-Empowered Semantic Communication Systems with a Shared Knowledge Base

Peng Yi,Yang Cao,Xin Kang,Ying-Chang Liang

DOI: https://doi.org/10.1109/TWC.2023.3330744

2023-11-06

Abstract:Deep learning-empowered semantic communication is regarded as a promising candidate for future 6G networks. Although existing semantic communication systems have achieved superior performance compared to traditional methods, the end-to-end architecture adopted by most semantic communication systems is regarded as a black box, leading to the lack of explainability. To tackle this issue, in this paper, a novel semantic communication system with a shared knowledge base is proposed for text transmissions. Specifically, a textual knowledge base constructed by inherently readable sentences is introduced into our system. With the aid of the shared knowledge base, the proposed system integrates the message and corresponding knowledge from the shared knowledge base to obtain the residual information, which enables the system to transmit fewer symbols without semantic performance degradation. In order to make the proposed system more reliable, the semantic self-information and the source entropy are mathematically defined based on the knowledge base. Furthermore, the knowledge base construction algorithm is developed based on a similarity-comparison method, in which a pre-configured threshold can be leveraged to control the size of the knowledge base. Moreover, the simulation results have demonstrated that the proposed approach outperforms existing baseline methods in terms of transmitted data size and sentence similarity.

Artificial Intelligence,Information Theory,Machine Learning

What problem does this paper attempt to address?

The paper attempts to address two main issues:

1. **Improving the interpretability of semantic communication systems**: Most existing semantic communication systems adopt an end-to-end architecture. Although this architecture performs well, it is considered a black-box system and lacks interpretability. The paper proposes a new semantic communication system that introduces a shared knowledge base, making the system more transparent and interpretable.

2. **Enhancing semantic information theory**: Existing semantic communication systems lack support from information theory in their technical implementation, especially when defining semantic entropy. The paper provides a mathematical explanation based on a shared knowledge base, defining semantic self-information and source semantic entropy, thereby providing a theoretical foundation for the reliability and performance of semantic communication systems.

Specifically, the main contributions of the paper include:

1. **Proposing a new semantic communication framework**: This framework includes three participants—the sender, the receiver, and the shared knowledge base. The shared knowledge base consists of highly readable sentences, making the system more interpretable than existing end-to-end architectures.

2. **Providing a mathematical explanation of the system**: By introducing the shared knowledge base, the paper defines the semantic self-information of messages and source semantic entropy, enhancing the system's reliability.

3. **Proposing an effective knowledge base construction algorithm**: This algorithm is based on sentence similarity measurement and controls the size and remaining information of the knowledge base through a pre-configured threshold.

4. **Implementing a semantic communication system based on the self-attention mechanism**: Numerical simulation results show that this system outperforms other benchmark methods in terms of data transmission volume and sentence similarity.

In summary, the paper aims to improve the interpretability and theoretical support of semantic communication systems by introducing a shared knowledge base, thereby providing better performance and reliability in practical applications.