No label defined (Q4396398)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 1165689
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | No label defined |
scientific article; zbMATH DE number 1165689 |
Statements
14 June 1998
0 references
damping factor
0 references
signal filtering
0 references
Prony method
0 references
Cramer-Rao bound
0 references
0.8692448735237122
0 references
0.7623142600059509
0 references
The authors consider the noise-corrupted signal \(v_n=u_n+\varepsilon_n, n=0,1,\dots,N-1,\) where the signal \(u_n\) is the sum of real-valued quasipolynomials \[ u_n=\sum_{l=1}^L\sum_{k=1}^{K_l} e^{\lambda_l n}(a_{lk}n^{k-1}\sin(\omega_l n+\varphi_{lk})) \] with the damping factors \(\lambda_l<0\), frequencies \(\omega_l>0\), amplitudes \(a_{lk}\in R\), phases \(\varphi_{lk}\) and \(\varepsilon_n\) is a real-valued white Gaussian noise with zero mean and variance \(\sigma^2\). The problem is to estimate the damping factors \(\lambda_l<0\) and frequencies \(\omega_l>0\) from the noisy data \(v_0,v_1,\dots,v_{N-1}\). A two-stage method is proposed to examine the problem. In the first stage the Prony method with prefiltering is used to obtain initial estimates. The second stage incorporates steps of updating of the initial estimates with the help of an iterative optimization procedure. The algorithm is investigated by means of numerical simulations.
0 references