Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
Conventionally, if equalization is carried out in time domain, matrix inversion is involved. However, if the channel has long impulse response, the computational complexity of calculation matrix inversion is very high, making it impossible for practical applications. We recall that the received...
Simply put, the MMSE equalizer coefficients are given as w(k,t)=H*(k,t)/(sqr(|H(k,t)|)+1/SNR).
As SNR increases, MMSE criterion converges to zero-forcing one. In the above equation, H(k,t) is the DFT of h. So, you have to take the DFT of channel gains.
hope that helps
Good explanation. We should emphasize again that the wise-product of two DFT relates to the circular convolution of the two corresponding time-domain sequences. Zero-oadding is thus required.
I get confused with you statement. It comes to me that you perhaps perform the FFT of the input signal via the command "fft" in Matlab. If such a case, you can use the "ifft" to get back the input signal.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
