A testing approach for a configurable RMPI-based Analog-to-Information Converter

Abstract

Compressive Sensing (CS) has recently received a lot of attention due to the fact that it takes advantage from the sparse representation of some natural signals achieving acquisition and energy efficient by using Analog-to-Information Converter (AIC). In short, an AIC receives an analog input signal and outputs a digitized and compressed of this signal. Even though there are several AIC architectures, there is still a lack of testing approaches for these converters. In this paper, a testing method is adapted and applied to evaluate a physical implementation of a configurable architecture AIC based on the Random Modulation Pre-Integrator (RMPI). By means of this testing method, it can be obtained the Signal-to-Noise and Distortion Ratio (SINAD) directly from the AIC output without using signal reconstruction algorithm. This test method has been validated for this architecture by simulations and by means of experimentation in hardware. Results show that some features can be extracted by means of SINAD, such as the best range of input signal amplitude to be used and which configurable parameters best fit the converter.