Robust Nonlinear Transform Coding: A Framework for Generalizable Joint Source-Channel Coding

2025-11-24

AI Summary

To address the poor generalization of digital joint source-channel coding (JSCC) under dynamic channels and its reliance on channel-specific training, this paper proposes Robust Nonlinear Transform Coding (Robust-NTC). Methodologically, Robust-NTC adopts variational latent-variable modeling and introduces, for the first time, a Gaussian surrogate variational objective to explicitly characterize the latent-space distribution and decouple channel perturbations. It integrates element-wise quantization, adaptive bit-depth selection, and OFDM system constraints within a unified resource allocation framework, jointly optimizing quantization, modulation order, and power allocation. Experimental results demonstrate that Robust-NTC achieves significantly improved rate-distortion performance and stable reconstruction fidelity across a wide SNR range—without requiring channel-specific training—while simultaneously reducing OFDM transmission latency.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Search and Optimization: Mixed Discrete/Continuous Search
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

This paper proposes robust nonlinear transform coding (Robust-NTC), a generalizable digital joint source-channel coding (JSCC) framework that couples variational latent modeling with channel adaptive transmission. Unlike learning-based JSCC methods that implicitly absorb channel variations, Robust-NTC explicitly models element-wise latent distributions via a variational objective with a Gaussian proxy for quantization and channel noise, allowing encoder-decoder to capture latent uncertainty without channel-specific training. Using the learned statistics, Robust-NTC also facilitates rate-distortion optimization to adaptively select element-wise quantizers and bit depths according to online channel condition. To support practical deployment, Robust-NTC is integrated into an orthogonal frequency-division multiplexing (OFDM) system, where a unified resource allocation framework jointly optimizes latent quantization, bit allocation, modulation order, and power allocation to minimize transmission latency while guaranteeing learned distortion targets. Simulation results demonstrate that for practical OFDM systems, Robust-NTC achieves superior rate-distortion efficiency and stable reconstruction fidelity compared to digital JSCC baselines across wide-ranging SNR conditions.

Problem

• Develops a generalizable joint source-channel coding framework using variational latent modeling
• Enables adaptive quantization and bit allocation based on online channel conditions
• Optimizes transmission resources to minimize latency while maintaining reconstruction quality

Innovation

• Robust-NTC couples variational latent modeling with channel adaptation
• It explicitly models latent distributions using Gaussian proxy
• Unified resource allocation optimizes quantization and transmission parameters