Forums

who has matlab implementation of multistage weiner filter?

Started by jjdoc August 1, 2008
I've  read J. Scott Goldstein's  paper 'A Multistage Representation of
the Wiener
Filter Based on Orthogonal Projections'  and written a test matlab
program. But it
seems wrong. Who has ever implemented it ?
On Aug 1, 3:00 pm, jjdoc <tsaitow...@sina.com> wrote:
> I've read J. Scott Goldstein's paper 'A Multistage Representation of > the Wiener > Filter Based on Orthogonal Projections' and written a test matlab > program. But it > seems wrong. Who has ever implemented it ?
Full referene pls... K
On Aug 1, 2:53&#2013266080;pm, kronec...@yahoo.co.uk wrote:
> On Aug 1, 3:00 pm, jjdoc <tsaitow...@sina.com> wrote: > > > I've &#2013266080;read J. Scott Goldstein's &#2013266080;paper 'A Multistage Representation of > > the Wiener > > Filter Based on Orthogonal Projections' &#2013266080;and written a test matlab > > program. But it > > seems wrong. Who has ever implemented it ? > > Full referene pls... > > K
The paper is Goldstein J Scott, Reed Irving S, Scharf Louis L. 'A multistage representation of the wiener filter based on orthogonal p rojections'. IEEE Transactions on Information Theory, Vol.44 Nov. 1998:pp2943-2959. Who can help me to implement it with matlab? Thanks!
On Jul 31, 10:00&#2013266080;pm, jjdoc <tsaitow...@sina.com> wrote:
> I've &#2013266080;read J. Scott Goldstein's &#2013266080;paper 'A Multistage Representation of > the Wiener > Filter Based on Orthogonal Projections' &#2013266080;and written a test matlab > program. But it > seems wrong. Who has ever implemented it ?
Rather than asking for the solution, why don't you ask a more meaningful question, such as what you expect and what you get. Right now you come across as somebody asking for the solution for free, without having tried.
On Aug 2, 2:57&#2013266080;am, julius <juli...@gmail.com> wrote:
> On Jul 31, 10:00&#2013266080;pm, jjdoc <tsaitow...@sina.com> wrote: > > > I've &#2013266080;read J. Scott Goldstein's &#2013266080;paper 'A Multistage Representation of > > the Wiener > > Filter Based on Orthogonal Projections' &#2013266080;and written a test matlab > > program. But it > > seems wrong. Who has ever implemented it ? > > Rather than asking for the solution, why don't you ask > a more meaningful question, such as what you expect > and what you get. > > Right now you come across as somebody asking for > the solution for free, without having tried.
I'm so sorry to ask the question like that. Now, I've solve this problem myself. My original purpose is to adaptively estimate the frequecies from noisy harmonic signal. The subspace methods are seems to be good solutions, such as MUSIC , ESPRIT and so on. In these methods, however, the inverse of convariance matrix or the eigenvector of date vectors need to be calculated to get the noise subspace and the signal subspace. The computational complexity can be reduced by Multistage Wiener Filter. But at beginning, I can't understand the algorithm in J. Scott Goldstein's paper. From the Michael D. Zoltowski tutorial (http://cobweb.ecn.purdue.edu/~mikedz/research/ research.html), I got the algorithm's detail information. Now I'v implemented it with Matlab. Thank julius for his suggestion!