Answers are grounded in this paper and cite their sources.
Data Compression is a staple of data processing and storage. Sending and storing data more efficiently is an open challenge in the Internet-of-Things (IoT), with devices typically characterized by limited availability of energy and computing power. The problem tackled in this paper is the massive amounts of sensor data collected and sent uncompressed by IoT-devices. We address this issue by compressing local data using a neural network supplemented with the Residual Vector Quantization (RVQ) technique. This paper, inspired by lossy neural compressors for audio like Google Soundstream and Meta EnCodec, proposes EdgeCodec: a lightweight lossy neural compressor specifically designed to run at the edge on low-power and resource constrained Microcontroller Units (MCUs). EdgeCodec processes multi-channel data with a flexible end-to-end learnable pipeline. We evaluate EdgeCodec in a real-life challenging use case, namely wind turbine monitoring using a 40-channel barometric sensor. Under the proposed use-case, our EdgeCodec reaches a Compression Ratio (CR) between 2560 and 10240 that can be varied in real-time to tune the trade-off between compression and reconstruction quality. Executed onboard a low-power MCU, EdgeCodec achieves an average reconstruction error of 2.54% over the whole validation set. It requires only 52.6ms to compress the 16kb generated in 1s. Final results demonstrate how EdgeCodec can reduce the energy consumption by 2.9x for the required wireless transmission.
No takes yet. Share an insight, caveat, or question.
Hodo et al. (2025) studied this question.