I need to check the loopback flowgraph you mentioned, but the header detection should work down to very low Eb/N0 values.
However, you mention a noise amplitude of 1.5. IIRC, that would yield an SNR below 0 dB, not 13 dB as you mention. Cheers, Martin On 19.12.2015 17:41, Mateusz wrote: > Hello, I have been evaluating the OFDM examples (at the beginning of my > Bachelor Thesis) > In ofdm_rx I have added noise source and file sinks to plot BER(SNR) in > matlab. I added a Tag Debug on the output. > The result is disappointing, I observe plenty of lost packets already at > the SNR of about 13dB and below (noise amplitude about 1.5 in an > unmodified ofdm_rx). > So are those examples useful, with such a performance, to utilize with > USRP and antennas? > Which part of a synchronization process fails, that causes missing > indexes of data packets (observed in a Tag Debug) in a noisy channel)? > I think it's at the loop of 'header recognition', so it's the 'coarse > frequency offset' that is too big for 'channel estimator'. > > I mainly want to know if it's only possible to use ofdm_rx with a low > noise, with SNR over about 13 dB. > > Mateusz Loch > > _______________________________________________ > Discuss-gnuradio mailing list > Discuss-gnuradio@gnu.org > https://lists.gnu.org/mailman/listinfo/discuss-gnuradio _______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
