Adapted Compressed Sensing for Effective Hardware Implementations by Mauro Mangia Fabio Pareschi Valerio Cambareri Riccardo Rovatti & Gianluca Setti

Author:Mauro Mangia, Fabio Pareschi, Valerio Cambareri, Riccardo Rovatti & Gianluca Setti
Language: eng
Format: epub
Publisher: Springer International Publishing, Cham

7. Analog-to-Information Conversion

Mauro Mangia1 , Fabio Pareschi2, Valerio Cambareri3, Riccardo Rovatti4 and Gianluca Setti2

(1)ARCES, Università di Bologna, Bologna, Italy

(2)ENDIF, Università di Ferrara, Ferrara, Italy

(3)ICTEAM/ELEN, Université Catholique de Louvain, Louvain-la-Neuve, Belgium

(4)DEI, ARCES, Università di Bologna, Bologna, Italy

No monolithic implementations of Compressed Sensing (CS)-based acquisition systems have been proposed so far as commercial products. Yet, a number of prototypes have appeared in the scientific literature. In this chapter we present some hardware architectures recently proposed in the most important microelectronics conferences and journals, capable of working as CS-based analog-to-information converters (AICs).

All the works considered here share a common methodology. In fact, despite being very different from each other, the front-end implementation is always analog, i.e., all of them belong to the class of analog Compressed Sensing systems exploiting the random modulation pre-integration (RMPI) architecture. The order in which these works are considered is the same order in which they appeared in the literature.

Conversely, no architectures exploiting the random sampling (RS) architecture is considered here, even if some implementation of AIC based on this approach has been proposed so far [26]. The main reason is that a RS-based AIC is basically a standard analog-to-digital converter (ADC) where randomization is added into the control logic. However, the core architecture is typically the very same of a standard ADC with performance (bandwidth, precision, etc.) similar to that of the ADC embedded in the AIC, while RMPI solutions require a full-custom design. Furthermore, based on the observations of Chap. 6, RS architectures are not general purpose ones and are able to achieve good performance only on a limited class of input signals.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.