UDSem: A Unified Distributed Learning Framework for Semantic Communications Over Wireless Networks

UDSem: унифицированная архитектура распределённого обучения для семантических коммуникаций в беспроводных сетях
Ping Zhang, Xiaodong Xu, Guoshun Nan, Xinchen Lyu, Qimei Cui, Xinghan Liu
2023-04-24

distributed semantic communicationsfederated learningmodular neural network trainingsemantic interpretationwireless networks
End-to-end semantic communications (ESC) rely on deep neural networks (DNN) to boost the communication efficiency by only transmitting the semantics of data. However, ESC is data-hungry for training millions of neural parameters, while uploading massive data generated by users/devices to a central server is usually impractical due to privacy issues, regulatory guidelines and network congestion, etc. Inspired by federated learning, where each device contributes to the parameters' update by independently computing the gradient based on its local data, we present UDSem, a unified distributed learning framework of semantic communications for texts and images over wireless networks. The key ingredients of our UDSem include 1) a flexible learning mechanism that can be tailored to a device based on its computing capabilities, 2) an efficient learning approach that splits the system into multiple modules and updates the parameters independently both on clients and the central server, 3) a mixed aggregation method that globally updates the neural parameters for each module of the ESC system. Equipped with these ingredients, our proposed UDSem is able to properly fill the gap between the training efficiency of ESC systems and the privacy requirement of users' data. Experimental results over two public benchmarks show the superiority of our UDSem in terms of convergence speed and semantic interpretation, potentially paving the way for distributed semantic communications in future wireless networks.
1
A mixed aggregation method globally updates neural parameters for each module while avoiding centralized collection of users’ raw training data.
2
Experiments on two public benchmarks show improved convergence speed and semantic interpretation compared with the evaluated alternatives.
3
The framework supports device-adaptive learning, tailoring training mechanisms to heterogeneous client computing capabilities.
4
UDSem decomposes the semantic communication system into multiple modules whose parameters are updated independently on clients and the central server.
5
UDSem is a unified distributed-learning framework for training end-to-end semantic communication systems for both text and image transmission over wireless networks.

end-to-end semantic communication systems for text and image transmission over wireless networks

distributed training efficiency, privacy preservation, convergence speed, and semantic interpretation of semantic communication neural networks

Publication Details
Publication Date
2023-04-24
Journal
Publisher
ISSN
Cited by
19
Access Type
Author Information
Authors
Ping Zhang
Xiaodong Xu
Guoshun Nan
Xinchen Lyu
Qimei Cui
Xinghan Liu
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat →
Make a presentation
100%