How up-sampling is implemented on digital data

[elec-eng]

Hi all,

I am currently implementing the PHY layer of a communication system. The transmitter has a pulse shaping filter followed by a DAC, as depicted below:



My question is:
How up-sampling is being implemented in hardware on input data bits, i.e. is it implemented by zero padding or by just repeating bits

for example, the pattern 1 0 1 , how will it look like after up-sampling if the OSR=8

would it be 10000000 00000000 10000000
or 11111111 00000000 11111111


Thanks

 

[zorx]

Re: How oversampling is implemented on digital data

If I understood correctly, what you need is an "up-sampling" operation. Oversampling is sampling of analogue signal with sampling frequency much higher than Nyquist frequency.
Up-sampling and down-sampling are multirate techniques that are used to convert the given sampling rate (on already sampled signal) to the desired sampling rate, and to provide different sampling rates through the system without destroying the signal components of interest.
On this subject there is a very good book:
"Multirate Filtering for Digital Signal Processing: MATLAB Applications" by Ljiljana Milic from University of Belgrade, Serbia.

As an answer to your question, this is a quote from that book:

"Up-Sampling Operation

The up-sampling by an integer factor L is performed by inserting L-1 zeros between two consecutive samples. Applying the up-sampling operation to the discrete signal {x[n]}, produces the up-sampled signal {y[m]} where



The up-sampling operation increases the sampling rate FT of the original signal x(nT). The sampling frequency FT’ of the signal y(mT’) is L times larger than the sampling rate of the original signal, i.e, FT’=L*FT.



"

Hope this helps a little.

I strongly suggest you to buy this book if you are going to work with multirate systems, especially because it comes with Matlab examples, and explains things in a very simple and straightforward way.

 

[elec-eng]

Thanks very much zorx

You understood what i meant correctly. Indeed, I meant up-sampling not oversampling.

I have edited my post.