Título: Non-Negative Matrix Factorization For Improving Passive Sonar Signal Detection
Autores: Moura, N. N. de; Paladino, Igor; Seixas, J. M. de
Resumo: Non-negative matrix factorization (NMF) has been shown to be useful for decomposition of multivariate data. In this paper, NMF will be implemented using the alpha divergence and obtain blind signal separation (BSS). The aim is to improve the signal/interference ratio of a passive sonar system that suffers from mutual interference in adjacent bearings and from the self-noise.
Palavras-chave: Passive Sonar; Spectral Analysis; DEMON Analysis; Blind Source Separation; Non-Negative Matrix Factorization; Divergence
Código DOI: 10.21528/CBIC2011-17.5
Artigo em pdf: st_17.5.pdf
Arquivo BibTex: st_17.5.bib