The Concept of a Measurement Data Acquisition Platform Based on Compressive Sensing

The paper introduces the concept of a modern software-hardware platform for data acquisition and analysis, capable of efficiently handling vast amounts of measurement data in real time with minimal energy consumption. The current methodologies for information acquisition are predicated upon traditional sampling techniques, which frequently yield redundant data necessitating subsequent compression. The novel approach is based on compressive sensing (CS) technology, which involves sampling the signal at a rate lower than the Nyquist frequency. The CS measurement protocol combines sampling and data compression simultaneously, leading to more efficient hardware implementation of signal acquisition systems. Different from commonly used methods, the innovative CS technique estimates the measurement signal using a small set of linear, incoherent measurements that are simple to store or transmit. CS is an effective way to reduce the required bandwidth and memory resources without sacrificing any information. In the CS acquisition strategy, random modulator pre-integration is applied. In the software part of the monitoring platform, the problem of signal parameters estimation, identification of spectral components or time-frequency spectrograms, direct inference, and direct spectrum detection is solved in two ways, with a reconstruction algorithm and using a reconstruction-free path.