Semantic Communication Meets Edge Intelligence: Semantic-Relay-Aided Text Transmissions

Zeyang Hu1; Changsheng You1; Tianyu Liu1; Dingzhu Wen2; Ye Hu3; Yuanhao Cui1; Yi Gong1; Kaibin Huang4

2024

10.1109/JIOT.2024.3433431

IEEE Internet of Things Journal   影响因子和分区

2372-2541

Semantic communication (SemCom) has emerged as a promising technology to improve the spectrum efficiency of next-generation wireless networks, by extracting meaningful content from data and transmitting relevant semantic information only. However, existing research usually overlooks the limited computing and storage resources on mobile devices, which may make it unaffordable to implement resource-demanding deep learning (DL)-based semantic encoders/decoders. Moreover, besides the end-to-end SemCom framework, cooperative SemCom has not been well studied in existing works, which can further enhance the communication performance. To address these issues, we propose a new architecture in this paper, called semantic relay (SemRelay), which acts as an edge server to provide DL-enabled SemCom (DeepSC) services for two categories of edge users, called semantic users (SemUsers) with rich computing resources and conventional users (ConUsers) with limited resources. Two new transmission protocols are proposed for enabling text transmissions from the BS to the SemUsers and ConUsers, respectively, via the SemRelay (edge server). Moreover, an optimization problem is formulated to jointly design the SemRelay transmit power allocation and system bandwidth allocation to maximize the weighted sum-rate of all users. Although this problem is non-convex and hence difficult to solve, we propose an efficient algorithm to obtain a high-quality suboptimal solution by applying the block coordinate descent and successive convex approximation techniques. Finally, numerical results demonstrate the effectiveness of our proposed algorithm and the superior performance of the proposed SemRelay as compared to traditional decode-and-forward relays, especially in the small bandwidth regime.

第一

1.Department of Electronic and Electrical Engineering, Southern University of Science and Technology, Shenzhen, China
2.School of Information Science and Technology, ShanghaiTech University, Shanghai, China
3.Department of Industrial and Systems Engineering, University of Miami, Coral Gables, FL, USA
4.Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong, Hong Kong

Zeyang Hu,Changsheng You,Tianyu Liu,et al. Semantic Communication Meets Edge Intelligence: Semantic-Relay-Aided Text Transmissions[J]. IEEE Internet of Things Journal,2024,PP(99).

Zeyang Hu.,Changsheng You.,Tianyu Liu.,Dingzhu Wen.,Ye Hu.,...&Kaibin Huang.(2024).Semantic Communication Meets Edge Intelligence: Semantic-Relay-Aided Text Transmissions.IEEE Internet of Things Journal,PP(99).

Zeyang Hu,et al."Semantic Communication Meets Edge Intelligence: Semantic-Relay-Aided Text Transmissions".IEEE Internet of Things Journal PP.99(2024).