The SCS Pactor modems use Viterbi encoding/decoding according to the information published by Dr. Rink in the "RTTY Digital Jounal" in the mid 1990's:
"Very efficient error control coding using a convolutional code with a constraint length of 9 and a real Viterbi decoder with soft decision is applied at all speed levels, in addition to analog Memory-ARQ." Although the original Pactor protocol was FSK and the P2 and P3 modems use PSK, the memory ARQ concept should be similar. SCS does claim that P3 is more robust than P2, but at the slowest speed needed for the worst conditions (2 tone) the two modems seem about the same. I very much would like to see some real world comparisons between the "ALE" 8PSK2400 modems and Pactor and other amateur modes as well. When I look at the actual claims by manufacturers (often graphs based upon theoretical computer simulations) the 8PSK2400 modes do not seem to work well much below zero dB and what is even more of a concern is that they sometimes consider 2 msecs of ISI as being a high degree of ISI. It is not that unusual to have 5 or more msecs of ISI on the lower HF bands, which I understand can prevent even 45 baud RTTY signals from working. The other claims tend to show how many messages can be sent over a time period between two points with different sun spot numbers and it is not unusual to show ZERO throughput without fairly high power levels. Government/commercial sites may use power levels well above what most of us use for digital modes, typically running at 25 watts or so with a 100 watt transmitter. 73, Rick, KV9U Jose A. Amador wrote: > Rick wrote: > > >> Hi Jose, >> >> Do you see any difference between the convolutional code of Pactor and >> the Viterbi code in MFSK16 or Patrick's use of Viterbi code in his F >> versions of PSK and his FEC of DEX? >> > > I have not studied those in detail. And I know about Viterbi DECODER, > but, would a Viterbi encoder be the same thing as a convolutional encoder? > > In the PTC-II manual, they write about a convolutional encoder with > constraint length of nine. But you also need to know how to take the > taps. That is, to achieve a Pactor II/III compatible decoder. Other > variants could be as good, but would not be compatible. > > >> The extra thing that SCS developed early with Pactor 1 was the ability >> to create a correct frame from multiple tries (what they call memory >> ARQ) which has a similiar effect as coding gain. >> > > Yes, but Pactor and P2/P3 are "different animals". Plain Pactor uses > FSK, P2 and P3 are variants of PSK. > > >> If it is ever proven that the wide bandwidth, high speed ALE modes using >> 8PSK can match P3, then those protocols could be used as an open source >> design, but when you look at the theoretical throughputs from various >> sources, they seem to suggest that these modes do not work all that well >> below zero dB S/N. >> > > Even P3 is affected by low SNR. It excels above + 10 dB or so. > > 8PSK is already in a disadventageous position compared to BPSK or QPSK. > It has a smaller Hamming distance, so, it is less robust. > > It is already clear in Carlson and Sklar books, and possibly in a dozen > or more engineering books dealing with digital modulations in > communications systems. > > Nevertheless, hats off to the SCS team. They knew what to do, and how to > code it as well. People can disagree on other aspects, but it is > undeniable that it has been a well done job. > > 73, > > Jose, CO2JA > > > > __________________________________________ > > Participe en Universidad 2008. > 11 al 15 de febrero del 2008. > Palacio de las Convenciones, Ciudad de la Habana, Cuba > http://www.universidad2008.cu > > > Announce your digital presence via our Interactive Sked Page at > http://www.obriensweb.com/drsked/drsked.php > > Yahoo! Groups Links > > > > > > >
