Vision Transformer for Adaptive Image Transmission over MIMO Channels

TL;DR

This paper introduces ViT-MIMO, a vision transformer-based joint source and channel coding scheme for wireless image transmission over MIMO channels, which adapts to different channel conditions without retraining and improves transmission quality.

Contribution

The paper proposes a novel ViT-based JSCC scheme for MIMO systems that adaptively learns feature mapping and power allocation, outperforming traditional separation-based methods.

Findings

Significant improvement in transmission quality across various scenarios.

Effective adaptation to different channel conditions without retraining.

Outperforms separation-based benchmarks.

Abstract

This paper presents a vision transformer (ViT) based joint source and channel coding (JSCC) scheme for wireless image transmission over multiple-input multiple-output (MIMO) systems, called ViT-MIMO. The proposed ViT-MIMO architecture, in addition to outperforming separation-based benchmarks, can flexibly adapt to different channel conditions without requiring retraining. Specifically, exploiting the self-attention mechanism of the ViT enables the proposed ViT-MIMO model to adaptively learn the feature mapping and power allocation based on the source image and channel conditions. Numerical experiments show that ViT-MIMO can significantly improve the transmission quality cross a large variety of scenarios, including varying channel conditions, making it an attractive solution for emerging semantic communication systems.