IEEE access, 2017-01, Vol.5, p.26460-26465
2017
Volltextzugriff (PDF)
Details
- Autor(en) / Beteiligte
- Titel
- Improved Singular Value Decomposition (TopSVD) for Source Number Estimation of Low SNR in Blind Source Separation
- Ist Teil von
- IEEE access, 2017-01, Vol.5, p.26460-26465
- Ort / Verlag
- Piscataway: IEEE
- Erscheinungsjahr
- 2017
- Quelle
- Open access e-journals list
- Beschreibungen/Notizen
- An improved singular value decomposition based on Toeplitz (TopSVD) is proposed to solve the problem of inaccurately estimating source numbers under the condition of a low signal-to-noise (SNR) ratio for blind source separation. First, Toeplitz modifies the covariance of the received data, and singular value decomposition is used to estimate the number of signal sources. The advantages of TopSVD over traditional approaches are demonstrated by simulated signals. The results demonstrate that the proposed method can be used to estimate the number of coherent sources under low SNR conditions; at the same time, it can significantly improve the accuracy of source number estimation under the conditions of a low SNR and coherent signal source with the simple algorithm.
- Sprache
- Englisch
- Identifikatoren
- ISSN: 2169-3536eISSN: 2169-3536DOI: 10.1109/ACCESS.2017.2754487Titel-ID: cdi_crossref_primary_10_1109_ACCESS_2017_2754487
- Format
- –
- Schlagworte
- Algorithm design and analysis, Algorithms, Blind source separation, Covariance, Covariance matrices, Decomposition, Eigenvalues and eigenfunctions, Estimation, low SNR, Matrix decomposition, Separation, Signal processing, Signal to noise ratio, Singular value decomposition, toeplitz