SNR and Resource Adaptive Deep JSCC for Distributed IoT Image Classification

2025-06-12

AI Summary

Existing joint source-channel coding (JSCC)-based deep neural network (DNN) splitting methods for wireless image classification—deployed across resource-constrained IoT devices and edge servers—rely on fixed network partitioning and static configurations, rendering them ill-suited to dynamic channel conditions and heterogeneous computational budgets. Method: We propose a dual-adaptive JSCC framework that jointly optimizes the CNN split point and coding structure in response to both signal-to-noise ratio (SNR) and floating-point operations (FLOPs) constraints. To enable efficient optimization under strict computational limits, we introduce the Learning-Assisted Intelligent Genetic Algorithm (LAIGA), which integrates random forests to prune infeasible configurations and embed domain-specific preferences. Contribution/Results: Evaluated under challenging conditions—ranging from −10 dB SNR to 1M–70M FLOPs—the framework achieves a 10% absolute accuracy gain over state-of-the-art SNR-adaptive multi-layer JSCC approaches, effectively overcoming the limitations of conventional fixed-splitting paradigms.

Technology Category

• Machine Learning: Hardware-aware ML
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

Sensor-based local inference at IoT devices faces severe computational limitations, often requiring data transmission over noisy wireless channels for server-side processing. To address this, split-network Deep Neural Network (DNN) based Joint Source-Channel Coding (JSCC) schemes are used to extract and transmit relevant features instead of raw data. However, most existing methods rely on fixed network splits and static configurations, lacking adaptability to varying computational budgets and channel conditions. In this paper, we propose a novel SNR- and computation-adaptive distributed CNN framework for wireless image classification across IoT devices and edge servers. We introduce a learning-assisted intelligent Genetic Algorithm (LAIGA) that efficiently explores the CNN hyperparameter space to optimize network configuration under given FLOPs constraints and given SNR. LAIGA intelligently discards the infeasible network configurations that exceed computational budget at IoT device. It also benefits from the Random Forests based learning assistance to avoid a thorough exploration of hyperparameter space and to induce application specific bias in candidate optimal configurations. Experimental results demonstrate that the proposed framework outperforms fixed-split architectures and existing SNR-adaptive methods, especially under low SNR and limited computational resources. We achieve a 10% increase in classification accuracy as compared to existing JSCC based SNR-adaptive multilayer framework at an SNR as low as -10dB across a range of available computational budget (1M to 70M FLOPs) at IoT device.

Problem

• Adapts JSCC to varying SNR and computational budgets
• Optimizes CNN hyperparameters under FLOPs and SNR constraints
• Improves classification accuracy in low SNR conditions

Innovation

• SNR-adaptive distributed CNN framework
• Learning-assisted intelligent Genetic Algorithm
• Random Forests based hyperparameter optimization