Thread Subject: Frequency band splitting

Hi All,
I m new in the use of Matlab. I have to a model the transmission of a wideband signal through a communication channel with limited bandwidth in Matlab. I m required to split an input wideband signal into low and high bands. The channel will pass the low band part automatically, but i have to downshift the highband so that it can go through the same channel. At the output the downshifted highband will be upsifted, then the two parts will be combined to reconstruct the original signal.
Please what sort of filter can I use or design in matlab to help solve the problem. Also is this related to multirate signal processing.
If it is related, why the need for downsampling(decimation) and upsampling( interpolation).
Thanks

"Bayus" wrote in message <j7d81m$mch$1@newscl01ah.mathworks.com>...
> Hi All,
> I m new in the use of Matlab. I have to a model the transmission of a wideband signal through a communication channel with limited bandwidth in Matlab. I m required to split an input wideband signal into low and high bands. The channel will pass the low band part automatically, but i have to downshift the highband so that it can go through the same channel. At the output the downshifted highband will be upsifted, then the two parts will be combined to reconstruct the original signal.
> Please what sort of filter can I use or design in matlab to help solve the problem. Also is this related to multirate signal processing.
> If it is related, why the need for downsampling(decimation) and upsampling( interpolation).
> Thanks

The Wavelet Toolbox implements an orthogonal or biorthogonal fiter bank. See wavedec for a multilevel decomposition, or dwt for a single level.

Wayne

Hello, thanks for your reply. I m not familiar with the wavelet stuff. I just want to design a low pass and high pass filters to separate the bands.
I need to know know the best choice of filters that will not distort the signal and that will allow signal reconstruction. Thanks
"Wayne King" wrote in message <j7d8o6$o89$1@newscl01ah.mathworks.com>...
> "Bayus" wrote in message <j7d81m$mch$1@newscl01ah.mathworks.com>...
> > Hi All,
> > I m new in the use of Matlab. I have to a model the transmission of a wideband signal through a communication channel with limited bandwidth in Matlab. I m required to split an input wideband signal into low and high bands. The channel will pass the low band part automatically, but i have to downshift the highband so that it can go through the same channel. At the output the downshifted highband will be upsifted, then the two parts will be combined to reconstruct the original signal.
> > Please what sort of filter can I use or design in matlab to help solve the problem. Also is this related to multirate signal processing.
> > If it is related, why the need for downsampling(decimation) and upsampling( interpolation).
> > Thanks
>
> The Wavelet Toolbox implements an orthogonal or biorthogonal fiter bank. See wavedec for a multilevel decomposition, or dwt for a single level.
>
> Wayne

"Bayus" wrote in message <j7eda6$8d9$1@newscl01ah.mathworks.com>...
> Hello, thanks for your reply. I m not familiar with the wavelet stuff. I just want to design a low pass and high pass filters to separate the bands.
> I need to know know the best choice of filters that will not distort the signal and that will allow signal reconstruction. Thanks
> "Wayne King" wrote in message <j7d8o6$o89$1@newscl01ah.mathworks.com>...
> > "Bayus" wrote in message <j7d81m$mch$1@newscl01ah.mathworks.com>...
> > > Hi All,
> > > I m new in the use of Matlab. I have to a model the transmission of a wideband signal through a communication channel with limited bandwidth in Matlab. I m required to split an input wideband signal into low and high bands. The channel will pass the low band part automatically, but i have to downshift the highband so that it can go through the same channel. At the output the downshifted highband will be upsifted, then the two parts will be combined to reconstruct the original signal.
> > > Please what sort of filter can I use or design in matlab to help solve the problem. Also is this related to multirate signal processing.
> > > If it is related, why the need for downsampling(decimation) and upsampling( interpolation).
> > > Thanks
> >
> > The Wavelet Toolbox implements an orthogonal or biorthogonal fiter bank. See wavedec for a multilevel decomposition, or dwt for a single level.
> >
> > Wayne

Hi that is what a wavelet filter bank does. Try a longer wavelet, like 'sym8' or 'sym10'

Wayne