FSSC: Federated Learning of Transformer Neural Networks for Semantic Image Communication

TL;DR

This paper introduces a federated learning approach using Swin Transformer models for semantic image communication, improving privacy and performance in multi-user scenarios with significant PSNR gains.

Contribution

It proposes a novel federated learning framework for Swin Transformer-based semantic communication, enhancing privacy and outperforming traditional methods.

Findings

Outperforms typical JSCC algorithms in simulations

Global model increases PSNR by over 2dB after local semantics integration

Effective in multi-user semantic image communication scenarios

Abstract

In this paper, we address the problem of image semantic communication in a multi-user deployment scenario and propose a federated learning (FL) strategy for a Swin Transformer-based semantic communication system (FSSC). Firstly, we demonstrate that the adoption of a Swin Transformer for joint source-channel coding (JSCC) effectively extracts semantic information in the communication system. Next, the FL framework is introduced to collaboratively learn a global model by aggregating local model parameters, rather than directly sharing clients' data. This approach enhances user privacy protection and reduces the workload on the server or mobile edge. Simulation evaluations indicate that our method outperforms the typical JSCC algorithm and traditional separate-based communication algorithms. Particularly after integrating local semantics, the global aggregation model has further increased the Peak Signal-to-Noise Ratio (PSNR) by more than 2dB, thoroughly proving the effectiveness of our algorithm.