Re: LMMSE estimation
- From: "Hany" <hany.kamal@xxxxxxxxxxx>
- Date: 15 Jun 2006 01:54:13 -0700
Thanks alot Arthur for your reply.
But actually I'm following & implementing HIPERLAN/2 standard and hence
I'm using a preamble to estimate the channel .. I didn't got through
the tracking using pilots yet.
Thanks again
Hany
Arthur wrote:
You might want to try DFT based interpolation too, I have just
implemented it, seems to work fine
"DFT-based channel estimation in 2D-pilot-symbol-aided OFDM wireless
systems"
Fernandez-Getino Garcia, M.J. Paez-Borrallo, J.M. Zazo, S.
Vehicular Technology Conference, 2001. VTC 2001 Spring. IEEE VTS 53rd
Regards,
Arthur,
Hany wrote:
Hello ALL,
I'm trying to simulate a LMMSE equalizer for OFDM signal.
The general equation requires the calculation of (X.X^H) where X is
the transmitted signal OR >>-----> X is the diagonal matrix containing
the transmitted signal..!
And X^H is the hermitian of X (hermitian is the conjugate transpose of
the matrix)
I'm actually using the DL preamble to estimate the channel (as
HIPERLAN/2 says) so >>---> (X) would equal to this preamble vector..
Example : DL_Preamble = [1 -1 -1 1 -1 1 1 -1] "a vector of the
subcarriers"
then X = [1 -1 -1 1 -1 1 1 -1];
the diagonal matrix would be :
[1 0 0 0 0 0 0 0
0 -1 0 0 0 0 0 0
0 0 -1 0 0 0 0 0
0 0 0 1 0 0 0 0
0 0 0 0 -1 0 0 0
0 0 0 0 0 1 0 0
0 0 0 0 0 0 1 0
0 0 0 0 0 0 0 -1];
the matrix multiplication of this matrix and it's hermitian could not
be done on MATLAB .. it gives a matrix with "Inf" elements...!
Could anybody try this simple operation and help please...
How can I calculate this term in order to perform a LMMSE equalizer..
Thanks a lot
Hany
.
- Follow-Ups:
- Re: LMMSE estimation
- From: Arthur
- Re: LMMSE estimation
- References:
- LMMSE estimation
- From: Hany
- Re: LMMSE estimation
- From: Arthur
- LMMSE estimation
- Prev by Date: Re: LMMSE estimation
- Next by Date: Re: LMMSE estimation
- Previous by thread: Re: LMMSE estimation
- Next by thread: Re: LMMSE estimation
- Index(es):
Relevant Pages
|