High frequency OFDM

Hi,

I successfully implemented OFDM with two USRPs with LFRX/LFTX at
1-50MHz, connect by cable. Now I tried switching over to higher
frequencies via TVRX/WBX. While I receive a perfectly valid looking
spectrum with a good SNR, it does not seem to decode at >200MHz.
If I increase the sub-carrier spacing of the transmission and thus the
bandwidth, it works again. But why do I have to do that if the signal
and SNR look okay?

What destroys a OFMD signal at high frequencies? Is it phase noise? Is
the USRPs frequency resolution different at higher frequencies?

Any literature/papers on this topic would also be helpful - I am kind of
stuck here.

GMX DSL: Internet-, Telefon- und Handy-Flat ab 19,99 EUR/mtl.
Bis zu 150 EUR Startguthaben inklusive! Aktuelle Nachrichten aus Politik, Wirtschaft & Panorama | GMX

On Mon, Jun 28, 2010 at 7:37 AM, Joachim R. [email protected] wrote:

Hi,

I successfully implemented OFDM with two USRPs with LFRX/LFTX at 1-50MHz, connect by cable. Now I tried switching over to higher frequencies via TVRX/WBX. While I receive a perfectly valid looking spectrum with a good SNR, it does not seem to decode at >200MHz.
If I increase the sub-carrier spacing of the transmission and thus the bandwidth, it works again. But why do I have to do that if the signal and SNR look okay?

What destroys a OFMD signal at high frequencies? Is it phase noise? Is the USRPs frequency resolution different at higher frequencies?

Any literature/papers on this topic would also be helpful - I am kind of stuck here.

It’s probably due to the frequency offset. Since the local oscillator
has a certain offset at its nominal frequency, that offset is
multiplied when the signal is multiplied to a higher frequency. So the
higher the frequency, the more you are offset. In the case of the OFDM
signals, you are now farther off frequency than the lock-in range of
the receiver.

You already found one fix for this, which is making the subcarriers
larger. The other thing to do is look at the frequency offset between
the two devices and try to compensate for it.

There was a discussion about this a few months ago that you can look
for that explains a way to handle larger offsets better than the look
and tune method, which, while it works, isn’t a realistic solution.

Tom

What destroys a OFMD signal at high frequencies? Is it phase noise? Is
the receiver.

You already found one fix for this, which is making the subcarriers
larger. The other thing to do is look at the frequency offset between
the two devices and try to compensate for it.

There was a discussion about this a few months ago that you can look
for that explains a way to handle larger offsets better than the look
and tune method, which, while it works, isn’t a realistic solution.

Tom

Thanks for the reply!

Sadly I already noticed and compensated for the frequency offset you
mentioned and manually centered the frequency block that OFDM consists
of. I also adjusted the respective levels, so clipping should not be an
issue as well.

I just wonder why you can’t use my OFDM configuration (that works
perfectly at low frequencies) at a higher frequency. My two USRPs are
connected via cable and the SNR and signal spectrum look good, so it
should work, shouldn’t it? :frowning:


GRATIS für alle GMX-Mitglieder: Die maxdome Movie-FLAT!
Jetzt freischalten unter Aktuelle Nachrichten aus Politik, Wirtschaft & Panorama | GMX