Abstract
The canonical correlation analysis (CCA) of rational system identification is investigated for autoregressive moving-average (ARMA) spectral estimation at low SNR. The method is used to compute the parameters of the state-space Markovian model and its spectrum using CCA. It is shown that this approach yields significantly better results and improved resolution for low SNR. An interesting feature of the CCA is that the system parameters are sign symmetric, which reduces the computation cost by half. The performance of this method for spectral estimation of multiple sinusoids in noise is compared with singular value decomposition and the canonical vector method.
Original language | English (US) |
---|---|
Title of host publication | Proceedings - IEEE International Symposium on Circuits and Systems |
Editors | Anon |
Publisher | Publ by IEEE |
Pages | 1319-1322 |
Number of pages | 4 |
Volume | 2 |
State | Published - 1989 |
Externally published | Yes |
Event | IEEE International Symposium on Circuits and Systems 1989, the 22nd ISCAS. Part 1 - Portland, OR, USA Duration: May 8 1989 → May 11 1989 |
Other
Other | IEEE International Symposium on Circuits and Systems 1989, the 22nd ISCAS. Part 1 |
---|---|
City | Portland, OR, USA |
Period | 5/8/89 → 5/11/89 |
ASJC Scopus subject areas
- Electrical and Electronic Engineering
- Electronic, Optical and Magnetic Materials