[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Discuss-gnuradio] OFDM Training Sequence
From: |
Martin Braun |
Subject: |
Re: [Discuss-gnuradio] OFDM Training Sequence |
Date: |
Thu, 17 Mar 2016 09:12:03 -0700 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.5.1 |
Mateusz,
the sync words are both for the Schmidl & Cox synchronization. Also,
they're defined in the frequency domain, so you won't see the 'two
identical parts here'. Rather, you skip every 2nd carrier.
Cheers,
Martin
On 03/17/2016 04:47 AM, Mateusz wrote:
> Hey, I do my Bachelor Thesis on OFDM using GnuRadio.
>
> I am using the joint system created from tx_ofdm and rx_ofdm examples.
> In simulation I try to use a bigger fft. So far I have managed to send
> and receive for fft_len=128 and fft_len=256, BPSK modulation. I ask to
> help me understand the creation of sync_words for bigger fft.
>
> I see the influence of the training sequences which I use on the quality
> of transmission. Of course I have read that the synchronization is based
> on "PN Sequence Correlation" and "Maximal Likelihood" papers. I suppose
> that sync_word1 and sync_word2 are based on those algorithms
> respectively. However I don't see "two identical parts of PN sequence"
> like it was described in the Schmidl&Cox paper.
>
> I tried to use the sequences containing 0's where needed, and random
> -1's and 1's elsewhere, but then the timing synchronization error
> (missing trigger signal in sync_sc) occurs more often, than when I use
> the replicas of original sync_words from the OFDM example.
>
> I attach the file with the training sequences which I have used.
>
> Regards,
> Mateusz Loch
>
>
> _______________________________________________
> Discuss-gnuradio mailing list
> address@hidden
> https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
>