On compressed sensing and super-resolution in DFT-based spectral analysis

Matteo Bertocco, Guglielmo Frigo, Claudio Narduzzi

Abstract

The paper discusses a novel frequency interpolation and super-resolution method for multitone waveform analysis, where a compressive sensing algorithm is employed to process data. Each signal acquisition involves a short data record, whose DFT coefficients are computed. A set of compressed measurements is obtained by taking records with different known starting instants, and employed to determine, by solving an orthogonal matching pursuit problem, the set of frequency components of the analysed waveform. Interpolation is presented as a compressed sensing problem and algorithm performances discussed.

Keywords

spectral analysis, frequency interpolation, compressed sensing

Download

IMEKO-TC4-2013-143.pdf

DOI

-

IMEKO TC

TC4 - Measurement of Electrical Quantities

Back to the proceedings