Some good progress on support for VHF/UHF data using Codec 2 modems:

  http://www.rowetel.com/?p=7207

- David

On 22/6/20 7:24 am, David Rowe wrote:
> Hi Glen,
> 
> Sounds like you are quite experienced and have had some fun times with
> LDPC work in the past.  Thanks for your offer but I think we are good
> with LDPC in Codec 2.
> 
> We currently have support for LDPC at the Octave simulation and C level
> using the CML library with great support on from Bill, VK5DSP.  It's all
> running on the stm32 thanks to Don's fine work with the 700D port last year.
> 
> Cheers,
> David
> 
> On 21/6/20 9:19 pm, glen english wrote:
>> David, I have quite a bit of code in this area from modem stuff from
>> years ago.
>>
>> Mostly based on David Mackay's work from memory. I can dig it all out.
>>
>> I generated :
>>
>> LDPC matrix generator, parametizable.
>>
>> C++ encoder, C encoder, encoders with precalculated LUTs that I used on
>> STM32L0 at 16 MHz ..
>>
>> My decoder (soft) was all C/C++ for a processor with a bit of memory and
>> speed, mildly optimized.
>>
>> But, depending on the code size, and rate rate, might work in a STM32.
>>
>> It was all teamed up with a chirp + D^2-BPSK hybrid  modem with an
>> optimized decoder  ( and highly NEON optimized to run on an cortex A5
>> and keep up.
>>
>> all is in a linux environment. I can dig out the LDPC stuff if ya like.
>>
>> -glen
>>
>>
>>
>> On 13/06/2020 3:31 pm, Bill Cowley wrote:
>>> Hi Jeroen,
>>>
>>> Further to David's reply, you might like to look at the AFF3CT
>>> sitehttps://aff3ct.github.io/  which provides comprehensive info on
>>> channel coding.  David's codes have been Repeat Accumulate (RA) LDPC
>>> codes - we used them initially as encoding is very easy (eg well
>>> suited to a microController).   Of course you probably don't need this
>>> - so you might get slightly better performance with other options.   I
>>> haven't played with the AFF3CT software much, but it does cover RA.
>>>
>>> Cheers, Bill
>>>
>>>     Cc:
>>>     Bcc:
>>>     Date: Thu, 11 Jun 2020 23:30:58 +0200
>>>     Subject: [Freetel-codec2] Generating LDPC code for 6000 mode
>>>     Hi,
>>>
>>>     The experiments with mode 6000 are going pretty well.
>>>     Another local HAM (PE1ITR) even managed to work our repeater with
>>>     just a
>>>     HackRF.
>>>     Compared to analog FM it seems to perform pretty well.
>>>
>>>     But I would like to take it a bit further and try to introduce
>>>     some FEC.
>>>     Since the currently used codec2 mode only needs 3200bps there is
>>> still
>>>     plenty of room.
>>>     I was thinking about a rate 2/3 code (687 code bits for 458 data
>>>     bits),
>>>     but I am a bit puzzled with the ldpc octave code.
>>>     How exactly do I generate the needed matrix (and ofcourse the C and
>>>     header files)?
>>>
>>>     Regards,
>>>     Jeroen
>>>
>>>
>>>
>>>
>>>
>>>
>>>     ---------- Forwarded message ----------
>>>     From: David Rowe <[email protected] <mailto:[email protected]>>
>>>     To: [email protected]
>>>     <mailto:[email protected]>
>>>     Cc:
>>>     Bcc:
>>>     Date: Fri, 12 Jun 2020 07:29:34 +0930
>>>     Subject: Re: [Freetel-codec2] Generating LDPC code for 6000 mode
>>>     Hi Jeroen,
>>>
>>>     Hey that's great work with working the new mode on a HackRF, is that
>>>     about 50mW max?
>>>
>>>     Yes, FEC is a fine idea.
>>>
>>>     A friend has been kind enough to generate the codes for me,
>>>     however it's
>>>     not a process that is automated or straight forward.  Couple of
>>>     options:
>>>
>>>     1/ Use an existing code.
>>>     2/ Use a code with more data bits than you require, and fill them
>>> with
>>>     known bits at the encoder and decoder that you don't transmit.  This
>>>     effectively lowers the code rate a little.  I do that for FreeDV
>>> 2020.
>>>     3/ Peruse the CML library codes and see if they have one that fits
>>>     your
>>>     use case, e.g. ldpcut.m
>>>
>>>     There is an Octave utility to generate .[ch] files and run the
>>>     codes in C.
>>>
>>>     Cheers,
>>>     David
>>>
>>>     On 12/6/20 7:00 am, Jeroen Vreeken wrote:
>>>     > Hi,
>>>     >
>>>     > The experiments with mode 6000 are going pretty well.
>>>     > Another local HAM (PE1ITR) even managed to work our repeater
>>>     with just a
>>>     > HackRF.
>>>     > Compared to analog FM it seems to perform pretty well.
>>>     >
>>>     > But I would like to take it a bit further and try to introduce
>>>     some FEC.
>>>     > Since the currently used codec2 mode only needs 3200bps there is
>>>     still
>>>     > plenty of room.
>>>     > I was thinking about a rate 2/3 code (687 code bits for 458 data
>>>     bits),
>>>     > but I am a bit puzzled with the ldpc octave code.
>>>     > How exactly do I generate the needed matrix (and ofcourse the C and
>>>     > header files)?
>>>     >
>>>     > Regards,
>>>     > Jeroen
>>>     >
>>>     >
>>>     >
>>>     > _______________________________________________
>>>     > Freetel-codec2 mailing list
>>>     > [email protected]
>>>     <mailto:[email protected]>
>>>     > https://lists.sourceforge.net/lists/listinfo/freetel-codec2
>>>     >
>>>
>>>
>>>     _______________________________________________
>>>     Freetel-codec2 mailing list
>>>     [email protected]
>>>     <mailto:[email protected]>
>>>     https://lists.sourceforge.net/lists/listinfo/freetel-codec2
>>>
>>>
>>>
>>> _______________________________________________
>>> Freetel-codec2 mailing list
>>> [email protected]
>>> https://lists.sourceforge.net/lists/listinfo/freetel-codec2
>>
> 
> 
> _______________________________________________
> Freetel-codec2 mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/freetel-codec2
> 


_______________________________________________
Freetel-codec2 mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

Reply via email to