Re: [Discuss-gnuradio] plateau detectors

[Martin Braun]

On 11/22/2014 06:01 PM, Marcus Müller wrote:
> Hi zs,
> hm, there's basically three places (that I know of) that GR uses Schmidl
> & Cox:
> the ofdm_chanest Channel estimator, the original S&C-based
> ofdm_sync_pn.py and the more robustness-optimized version in
> ofdm_sync_pnac.py.
> I suspect you're referring to one of the two ofdm_sync_pn*.py , but they
> differ a bit in how they deal with the peaks. Thus:
> Which implementation are you referring to exactly?

He's probably referring to plateau_detector_fb.

> Generally, Schmidl & Cox [1] states that
> "The timing metric reaches a plateau [...] For the AWGN channel, there
> is a window with a length of the guard interval where the metric reaches
> a maximum, and the start of the frame can be taken to be anywhere within
> this window without a loss in the received  SNR."
> Thus, without having really read throu ofdm_sync_pn.py in depth, if
> using the original approach, you shouldn't even care where the plateau
> is converted to a 1.

This. Also, how do you detect plateaus? Have a look at the code. Finding
the middle is easy, but how do you know you're at the beginning?

M