WARP Project Forums - Wireless Open-Access Research Platform

linrobert · 2017-Jul-16 23:39:06

Hello everyone,

In my project I want to implement the SCO estimation and compensation in my Rx PHY, and as far as I know doing this in frequency domain is easier than time domain.

I already implement the coarse frequency estimation and compensation, why I want to implement this is because the CFO between my transmitter and receiver is too large(about 40ppm).

Now I have to do the SCO part, how should I implement this function?

To my knowledge it can use the phase of two STS/LTS/ or Pilot to estimate the SCO, am I right?

I need some suggestion to do the SCO estimation and compensation.

Hope you can help me, thanks a lot!!

Btw, I am using the 802.11 reference design.

chunter · 2017-Jul-17 09:27:09

This is a pretty good paper that describes how SFO manifests in the frequency domain and how to correct for it. We actually do perform SFO correction in our M-code implementation of an OFDM Tx/Rx PHY for WARPLab. It should be straightforward to apply similar processing to the real-time 802.11 Reference Design PHY.

Here is the relevant chunk of code. If you have two WARPv3 nodes it might be useful to get a WARPLab installation up and running so you can experiment with that example and see what it is doing before diving into the System Generator model of the 802.11 PHY.

linrobert · 2017-Jul-18 12:05:06

Dear Chunter,

Thank you for your reply!

I want to ask two questions.

1. What is the limit of phase error correction can correct the phase error? If the phase error is larger than 2*pi can the equalizer fix it?

2.Thanks for post the SFO code to me, but I don't know what is the function of some code

What is the function of line 375,379,382,383? Or is there have some comments for me to read?

Thanks a lot!!

murphpo · 2017-Jul-18 12:32:22

The OFDM example implements the standard scheme for estimating SFO from pilot tones. You'll need to study the literature (Chris linked to a good resource above) and play with the code to better understand how SFO is estimated and corrected.

WARP Project Forums - Wireless Open-Access Research Platform

#1 2017-Jul-16 23:39:06

Sampling Frequency Offset (SFO) or Sampling Clock Offset (SCO)

#2 2017-Jul-17 09:27:09

Re: Sampling Frequency Offset (SFO) or Sampling Clock Offset (SCO)

#3 2017-Jul-18 12:05:06

Re: Sampling Frequency Offset (SFO) or Sampling Clock Offset (SCO)

#4 2017-Jul-18 12:32:22

Re: Sampling Frequency Offset (SFO) or Sampling Clock Offset (SCO)

Board footer