The equivalent source method as a sparse signal reconstruction

Efren Fernandez Grande, Angeliki Xenaki

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

This study proposes an acoustic holography method for sound field reconstruction based on a point source model, which uses the Compressed Sensing (CS) framework to provide a sparse solution. Sparsity implies that the sound field can be represented by a minimal number of non-zero terms, point sources in the case of this model. Sparse solutions can be achieved by l-1 norm minimization, providing accurate reconstruction and robustness to noise, because favouring sparsity suppresses noisy components. The study addresses the influence of the ill-conditioning of the propagation matrix, which can be a challenge for inverse problems where the energy in the solution vector is much greater than the energy in the observations (particularly in acoustic near-fields). Finally, the study examines the case of spatially extended sources and problems where the sparsity condition is not certainly guaranteed.
Original languageEnglish
Title of host publicationProceedings of inter.noise 2015
Number of pages10
Publication date2015
Publication statusPublished - 2015
Event44th International Congress and Exposition on Noise Control Engineering - San Francisco , United States
Duration: 9 Aug 201512 Aug 2015
Conference number: 44

Conference

Conference44th International Congress and Exposition on Noise Control Engineering
Number44
Country/TerritoryUnited States
CitySan Francisco
Period09/08/201512/08/2015

Fingerprint

Dive into the research topics of 'The equivalent source method as a sparse signal reconstruction'. Together they form a unique fingerprint.

Cite this