Digital-SC: Digital Semantic Communication with Adaptive Network Split and Learned Non-Linear Quantization

TL;DR

This paper introduces a digital semantic communication system that leverages adaptive network splitting, trainable non-linear quantization, and structured pruning to efficiently transmit semantic features for image classification tasks.

Contribution

It proposes a novel digital semantic communication framework with adaptive splitting, a trainable non-linear quantization module, and a semantic learning loss to improve robustness and efficiency.

Findings

Enhanced semantic feature representation through non-linear quantization.

Reduced transmission dimension via structured pruning.

Improved classification accuracy demonstrated on CIFAR-10 and ImageNet.

Abstract

Semantic communication, an intelligent communication paradigm that aims to transmit useful information in the semantic domain, is facilitated by deep learning techniques. Robust semantic features can be learned and transmitted in an analog fashion, but it poses new challenges to hardware, protocol, and encryption. In this paper, we propose a digital semantic communication system, which consists of an encoding network deployed on a resource-limited device and a decoding network deployed at the edge. To acquire better semantic representation for digital transmission, a novel non-linear quantization module is proposed to efficiently quantize semantic features with trainable quantization levels. Additionally, structured pruning is incorporated to reduce the dimension of the transmitted features. We also introduce a semantic learning loss (SLL) function to reduce semantic error. To adapt to various channel conditions and inputs under constraints of communication and computing resources, a policy network is designed to adaptively choose the split point and the dimension of the transmitted semantic features. Experiments using the CIFAR-10 and ImageNet dataset for image classification are employed to evaluate the proposed digital semantic communication network, and ablation studies are conducted to assess the proposed quantization module, structured pruning and SLL.