Selective Depthwise Separable Convolution for Lightweight Joint Source-Channel Coding in Wireless Image Transmission

2026-04-24

AI Summary

This work addresses the high computational complexity of existing deep learning–based joint source-channel coding (JSCC) models for wireless image transmission, which hinders their deployment on resource-constrained edge devices. The authors propose a configurable lightweight JSCC framework that enables flexible trade-offs between complexity and performance by selectively replacing standard convolutions with depthwise separable convolutions at varying depths of the encoder and decoder, according to predefined ratios. For the first time, the study systematically investigates how the placement and proportion of such replacements affect reconstruction performance, revealing that substituting intermediate layers yields an optimal balance and exposes significant inter-layer redundancy. The proposed method substantially reduces model parameters with only marginal degradation in image reconstruction quality, offering multiple efficient deployment options tailored for edge devices.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Computer Vision: Learning & Optimization for CV
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Cloud, edge and content delivery systems for the Web
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Depthwise separable convolutional (DSConv) layers have been successfully applied to deep learning (DL)-based joint source-channel coding (JSCC) schemes to reduce computational complexity. However, a systematic investigation of the layerwise and ratio-wise replacement of standard convolutional (Conv) layers with DSConv layers in JSCC systems for wireless image transmission remains largely unexplored. In this letter, we propose a configurable lightweight JSCC framework that incorporates a selective replacement strategy, enabling flexible substitution of standard Conv layers with DSConv layers at various layer positions and replacement ratios. By adjusting the proportion of layers replaced, we achieve different model compression levels and analyze their impact on reconstruction performance. Furthermore, we investigate how replacements at different encoder and decoder depths influence reconstruction quality under a fixed replacement ratio. Our results show that Conv-to-DSConv replacement at intermediate layers achieves a favorable complexity-performance trade-off, revealing layer-wise redundancy in DL-based JSCC systems. Extensive experiments further demonstrate that the proposed framework achieves substantial parameter reduction with only slight performance degradation, enabling flexible complexity-performance trade-offs for resource-constrained edge devices.

Problem

• joint source-channel coding
• depthwise separable convolution
• wireless image transmission
• model compression
• complexity-performance trade-off

Innovation

• selective replacement
• depthwise separable convolution
• joint source-channel coding
• lightweight neural network
• complexity-performance trade-off