Forwarded for your consideration:<br><br>---------- Forwarded message ----------<br><span class="gmail_quote">From: <b class="gmail_sendername">Larry Gadallah</b> <<a href="mailto:lgadallah@gmail.com">lgadallah@gmail.com
</a>><br>Date: Feb 26, 2007 9:48 PM<br>Subject: Re: Call for comments on open protocols<br>To: All things digital and fun <<a href="mailto:seatcp@wetnet.net">seatcp@wetnet.net</a>><br>Cc: <a href="mailto:linlink@wetnet.net">
linlink@wetnet.net</a><br><br></span>I agree with Ken that Phil Karn has written about many of the pertient<br>issues a long time ago. Johan Forrer's website<br>(<a href="http://www.johanforrer.net/">http://www.johanforrer.net/
</a>) also has a lot of information on HF<br>datacomms, including links to a GPL'd HF channel simulator.<br><br>So far, we've figured out that a new protocol should have the<br>following attributes:<br><br>1. Software defined solution, as much as possible
<br>2. True transparent data transfer, unrestricted alphabet (8 bit binary)<br>3. Scalable solution, able to use more or less bandwidth as available<br>4. Physical/Link layer protocol only, independent of network layer<br>
5. Adaptive link layer<br>6. Obey the end-to-end principle (<a href="http://www.ka9q.net/library.html">http://www.ka9q.net/library.html</a>)<br>7. Platform independent<br>8. Soft decision decoder (i.e. Viterbi, Fano, Turbo)
<br><br>On 2/26/07, Ken Koster <<a href="mailto:n7ipb@wetnet.net">n7ipb@wetnet.net</a>> wrote:<br>> This won't be new to many here but Phil Karn's proposal from somewhere<br>> around 1994 should be required reading.
<br>><br>> <a href="http://www.ka9q.net/dsp.html">http://www.ka9q.net/dsp.html</a><br>><br>> See the section titled 'New Link Layer Protocols' near the end.<br>><br>> In particular the discussion of convolutional codes and the ability to
<br>> provide for variable rate puncturing.<br>><br>> I'm not suggesting that Phil's proposal is what we should be going with<br>> (aside from the fact that it was never completed) but the techniques<br>
> he discusses are all relevant.<br>><br>> Ken, N7IPB<br>> EC/RO, Skagit County, WA<br>> ARES/RACES<br>><br>> _______________________________________________<br>> Seatcp mailing list<br>> <a href="mailto:Seatcp@wetnet.net">
Seatcp@wetnet.net</a><br>> <a href="http://wetnet.net/mailman/listinfo/seatcp">http://wetnet.net/mailman/listinfo/seatcp</a><br><br>-- <br>Larry Gadallah, VE6VQ/W7 lgadallah AT gmail DOT com<br>
PGP Sig: 616D 4E52 CF1F 3FEC FFFB F11B 7DB9 C79A EA7E B25B