Discussion:
[Freetel-codec2] Crest Factor Cohpsk
Steve
2017-02-03 05:23:29 UTC
Permalink
A long time ago I made a post about improving the crest factor by changing
to a bit-scrambled output (a la 9600 modems, etc). I was playing with
another idea today, and I'm pretty much convinced that bit-scrambling
doesn't improve anything.

Just looking at the spectrum shows the similarity, but then actually
measuring the peak and rms of the WAV files shows that not much is
happening.

Crest factor Clipping/Scrambled = 8.3491 dB (rms = 0.13624) (peak = 0.35626)
Crest factor Clipping/Raw = 8.3931 dB (rms = 0.13654) (peak =
0.35886)
.044 dB improvement - whoo-hoo...

I seem to remember a technical article about bit-scrambling being highly
effective. Alas, maybe not with a PSK waveform.

73
David Rowe
2017-02-03 05:57:52 UTC
Permalink
Hi Steve,

Well an interesting area to be working in - even if it didn't work out
this time. The 800XA waveform is interesting as it is a constant
amplitude mode (FSK modem) - if we can make it work on multipath channels.

I do recall a paper re using FEC coding to help with PAPR as well.

It's also worth trying tricks like offset qpsk or offsetting the
carriers in time wrt to each other. Even hard or soft amplitude
compression might help reduce PAPR, without impacting performance (ie
BER on a fading channel) much.

Cheers,

David
Post by Steve
A long time ago I made a post about improving the crest factor by
changing to a bit-scrambled output (a la 9600 modems, etc). I was
playing with another idea today, and I'm pretty much convinced that
bit-scrambling doesn't improve anything.
Just looking at the spectrum shows the similarity, but then actually
measuring the peak and rms of the WAV files shows that not much is
happening.
Crest factor Clipping/Scrambled = 8.3491 dB (rms = 0.13624) (peak = 0.35626)
Crest factor Clipping/Raw = 8.3931 dB (rms = 0.13654)
(peak = 0.35886)
.044 dB improvement - whoo-hoo...
I seem to remember a technical article about bit-scrambling being highly
effective. Alas, maybe not with a PSK waveform.
73
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
Loading...