Re: [Discuss-gnuradio] OFDM results.

[Andreas Müller]

> where is the inaccuracy in sampling frequency exactly happening?
> In USRP?

Yes - the sampling frequency is inaccurate, because the crystal on the
USRP is not oscillating at exactly 64 MHz.

FYI, the samples in the plot are actually not from an USRP (I didn't
record them myself) - the inaccuracy is about 74 PPM. With samples from
my USRP I found that the sampling frequency is off by about 14 PPM, and
in some samples that Jens Elsner recorded with his USRP I found an
inaccuracy of about 9 PPM. 

Depending on the number of subcarriers and constellation density, 9 PPM
can be enough to make the constellation a little more cloudy, if the
sample rate is not corrected.

Regards, Andreas

signature.asc
Description: OpenPGP digital signature