Premium
Stable sound source tracking based on two updating algorithms
Author(s) -
Owada Noboru,
Tsuji Daisuke,
Suyama Kenji
Publication year - 2011
Publication title -
ieej transactions on electrical and electronic engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.254
H-Index - 30
eISSN - 1931-4981
pISSN - 1931-4973
DOI - 10.1002/tee.20603
Subject(s) - algorithm , subspace topology , stability (learning theory) , tracking (education) , eigenvalues and eigenvectors , computer science , projection (relational algebra) , minification , matrix (chemical analysis) , signal subspace , qr decomposition , signal (programming language) , mathematical optimization , mathematics , noise (video) , artificial intelligence , machine learning , materials science , image (mathematics) , quantum mechanics , composite material , programming language , psychology , pedagogy , physics
In this paper, we present a stable sound source tracking method based on two updating algorithm. In the method, an eigenvector spanning the signal subspace is updated by using the Projection Approximation Subspace Tracking (PAST) algorithm without the eigen‐decomposition of a correlation matrix. Then, a constrained minimization problem is formulated taking into account the stability, and the solution is determined successively by applying the Interior Point Least Square (IPLS) algorithm. As a result, stable tracking can be achieved without the peak search of a MUltiple SIgnal Classification (MUSIC) spectrum, which often requires enormous computational costs because many complex multiplying operations are involved. Several experimental results in real room environments have demonstrated the high accuracy and the low computational costs of the proposed method. © Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.