Transient multiexponential data analysis using a combination of ARMA and ECD methods
Another attempt at estimating the time constants and number of components of multiple exponentials in white Gaussian noise is presented. Based on classical Gardner transform, the approach consists of two techniques. First, exponential compensation deconvolution method is used to deconvolved the d...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IACSIT Press, Singapore
2012
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/25573/ http://irep.iium.edu.my/25573/ http://irep.iium.edu.my/25573/1/029-ICCET2012-T10059.pdf |
| Summary: | Another attempt at estimating the time constants and number of components of multiple
exponentials in white Gaussian noise is presented. Based on classical Gardner transform, the approach
consists of two techniques. First, exponential compensation deconvolution method is used to deconvolved the
discrete convolution model arising from the application of Gardner transform. The deconvolved data is then
truncated and further processed using autoregressive moving average (ARMA) model whose AR parameters
are determined by using high-order Yule-Walker equations via the singular value decomposition (SVD)
algorithm. Simulations carried out using a number of synthetic signals demonstrate the effectiveness of the
proposed technique. Simulation results shows that this combination is more effective than many existing
techniques. It is clearly demonstrated that the proposed approach supersedes a number of popular techniques.
Its limitations are also highlighted. |
|---|