Semantics-Empowered Communication: A Tutorial-cum-Survey

Zhilin Lu,Rongpeng Li,Kun Lu,Xianfu Chen,Ekram Hossain,Zhifeng Zhao,Honggang Zhang

2023-11-11

Abstract:Along with the springing up of the semantics-empowered communication (SemCom) research, it is now witnessing an unprecedentedly growing interest towards a wide range of aspects (e.g., theories, applications, metrics and implementations) in both academia and industry. In this work, we primarily aim to provide a comprehensive survey on both the background and research taxonomy, as well as a detailed technical tutorial. Specifically, we start by reviewing the literature and answering the "what" and "why" questions in semantic transmissions. Afterwards, we present the ecosystems of SemCom, including history, theories, metrics, datasets and toolkits, on top of which the taxonomy for research directions is presented. Furthermore, we propose to categorize the critical enabling techniques by explicit and implicit reasoning-based methods, and elaborate on how they evolve and contribute to modern content & channel semantics-empowered communications. Besides reviewing and summarizing the latest efforts in SemCom, we discuss the relations with other communication levels (e.g., conventional communications) from a holistic and unified viewpoint. Subsequently, in order to facilitate future developments and industrial applications, we also highlight advanced practical techniques for boosting semantic accuracy, robustness, and large-scale scalability, just to mention a few. Finally, we discuss the technical challenges that shed light on future research opportunities.

Human-Computer Interaction,Artificial Intelligence,Networking and Internet Architecture,Signal Processing

What problem does this paper attempt to address?

The problems that this paper attempts to solve are as follows: With the rise of semantic - enhanced communication (SemCom) research, how to systematically sort out and summarize the background, research classifications, and technical details in this field, and provide a comprehensive technical tutorial to promote the understanding and development of semantic - enhanced communication in academia and industry.

Specifically, the paper aims to:

1. **Answer the "what" and "why" questions**: Explain the basic concepts of semantic - enhanced communication and its importance, and explore the differences between it and traditional communication methods.
2. **Construct the ecosystem of semantic - enhanced communication**: Including historical background, theoretical basis, measurement standards, data sets, and toolkits, etc.
3. **Propose research classifications**: Classify key enabling technologies according to explicit reasoning and implicit reasoning methods, and explain in detail how they promote the development of modern content and channel semantic - enhanced communication.
4. **Review and summarize the latest progress**: Discuss the relationship between semantic - enhanced communication and other communication levels (such as traditional communication) from an overall and unified perspective.
5. **Emphasize practical technologies and future challenges**: Introduce advanced practical technologies for improving semantic accuracy, robustness, and large - scale scalability, and discuss future research opportunities and technical challenges.

Through these efforts, the paper provides researchers and practitioners with a clear and in - depth perspective to help them better understand and promote the research and development of semantic - enhanced communication.

### Key Formulas Involved

Some of the key formulas involved in the paper are as follows:

- **Channel Model**:
\[
y = Hx+\varepsilon
\]
where \(y\) is the received signal, \(H\) is the channel gain matrix, \(x\) is the transmitted signal, and \(\varepsilon\) is the channel noise.

- **Semantic Information Quantity**:
\[
I(S;R)=H(S)-H(S|R)
\]
where \(I(S;R)\) represents the mutual information between the source \(S\) and the receiver \(R\), \(H(S)\) is the entropy of the source, and \(H(S|R)\) is the entropy under the condition of the given receiver.

- **Semantic Channel Capacity**:
\[
C_s=\max_{p(x)}I(X;Y)
\]
where \(C_s\) is the semantic channel capacity, \(p(x)\) is the probability distribution of the transmitted signal, and \(I(X;Y)\) is the mutual information between the transmitted signal \(X\) and the received signal \(Y\).

These formulas are used to describe the channel characteristics, information transmission efficiency, and the quantification of semantic information in semantic - enhanced communication.