Blind noncircular source separation in frequency domain
Hefa Zhang, Liping Li
Abstract
Hefa Zhang, Liping Li
Abstract
Independent component analysis (ICA) algorithms, ones of the most popular methods to solve the blind source separation (BSS) problems, can be classified into two main categories: time-domain and frequency-domain algorithms. The complex fast independent component analysis (c-FastICA) algorithm is one of the most popular methods for solving the ICA problems with complex-valued data. Novey and Adali extend the Complex FastICA algorithm to noncircular sources, named noncircular FastICA (nc-FastICA). In this paper, we solve the blind non-circular source separation (BNSS) of linear instantaneous mixture problem in frequency domain, by deriving a new fixed-point algorithm witch uses the noncircular information of pseudo-covariance matrix in frequency domain. Simulations are presented to demonstrate the effectiveness of our proposed method.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Independent component analysis (ICA) algorithms, ones of the most popular methods to solve the blind source separation (BSS) problems, can be classified into two main categories: time-domain and frequency-domain algorithms. The complex fast independent component analysis (c-FastICA) algorithm is one of the most popular methods for solving the ICA problems with complex-valued data. Novey and Adali extend the Complex FastICA algorithm to noncircular sources, named noncircular FastICA (nc-FastICA). In this paper, we solve the blind non-circular source separation (BNSS) of linear instantaneous mixture problem in frequency domain, by deriving a new fixed-point algorithm witch uses the noncircular information of pseudo-covariance matrix in frequency domain. Simulations are presented to demonstrate the effectiveness of our proposed method.
Key concepts: FastICA, Independent component analysis, Blind signal separation, Algorithm, Frequency domain, Computer science, Source separation, Covariance matrix