EdgeCodec: Onboard Lightweight High Fidelity Neural Compressor with Residual Vector Quantization

2025-07-08

AI Summary

To address the high energy consumption and limited sensor lifetime associated with wireless transmission of aerodynamic pressure sensing data from wind turbine blades, this paper proposes a lightweight, high-fidelity neural compression method tailored for embedded platforms. Methodologically, we design a highly asymmetric residual vector quantization (RVQ) autoencoder, integrated with adversarial training and per-sample dynamic bit-rate control, and perform end-to-end lightweight deployment and real-time inference optimization on the GAP9 microcontroller. Experimental results demonstrate compression ratios of 2560:1–10240:1, reconstruction error below 3%, and bit rates of only 11.25–45 bps; wireless transmission energy consumption is reduced by up to 2.9×, significantly extending edge sensor lifetime. To our knowledge, this is the first real-time neural compression implementation of RVQ combined with adversarial training on an ultra-low-power MCU, establishing a deployable, high-fidelity compression paradigm for edge sensing in rotating machinery.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Data Mining & Knowledge Management: Data Compression
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

We present EdgeCodec, an end-to-end neural compressor for barometric data collected from wind turbine blades. EdgeCodec leverages a heavily asymmetric autoencoder architecture, trained with a discriminator and enhanced by a Residual Vector Quantizer to maximize compression efficiency. It achieves compression rates between 2'560:1 and 10'240:1 while maintaining a reconstruction error below 3%, and operates in real time on the GAP9 microcontroller with bitrates ranging from 11.25 to 45 bits per second. Bitrates can be selected on a sample-by-sample basis, enabling on-the-fly adaptation to varying network conditions. In its highest compression mode, EdgeCodec reduces the energy consumption of wireless data transmission by up to 2.9x, significantly extending the operational lifetime of deployed sensor units.

Problem

• Compress wind turbine blade barometric data efficiently
• Achieve high compression rates with low reconstruction error
• Reduce energy consumption for wireless data transmission

Innovation

• Heavily asymmetric autoencoder for compression
• Residual Vector Quantizer enhances efficiency
• Real-time operation on GAP9 microcontroller