One other comment.

I have said before on this list that I have seen and used to have data produced 
by SouthWest Research Institute here in San Antonio that shows the maximim 
probable data capability of a single PSK signal.

This study was done for the U.S. Government in research to find the best 
robust, 
medium throughput mode for a nation command alert system os some sort.  I 
suspect it had something to do with always being able to keep the President 
informed under the most trying conditions with some sort of broadcast system.

The upshot of all this as a limited discussion of a number of hams that were at 
the reporting session that the current MIL-STD modems could be improved on but 
that to obtain the desired throughput you would need more than the bandwidth 
associated with normal SSB transmitters.

While amateur radio main not want a 4 or 5 KHz signal and the throughput that 
the government wanted, I think that a compromise bandwidth, something between 
that of PSK31 and perhaps 1 KHz with OFDM signal might be adequate for hams use 
on HF.  As many have said before...if you REALLY want/need 100 error free copy, 
you are going to need an ARQ function and FEC.

The "trick" is finding just how much of you signal you are going to give to FEC 
vs user data and how hard do you want to enforce ARQ.

73,

Walt/K5YFW

Walt DuBose wrote:
> Rud Merriam wrote:
> 
>>After a comment off list from Demeter I checked the Pactor specifications.
>>It uses DBPSK or DQPSK. 
>>
>>Why do the reports about Pactor indicate it is more robust than the QEX
>>article would indicate? 
>>
>>
>>Rud Merriam K5RUD 
>>ARES AEC Montgomery County, TX
>>http://TheHamNetwork.net
>>
> 
> Rud,
> 
> If you go back to the DCC presentation of KN6KB of a few years back on his 
> new 
> software modem...he measured the robustness of Pactor, MT63 and several other 
> modes and Pactor wasn't that much more robust than MT63 at a -5 dB SNR.
> 
> If I invested a $K Buck or so in Pactor III and WinLink, I'd claim it was the 
> best thing since sliced bread...woudln't you?
> 
> 73,
> 
> Wa;t/K5YFW
> 
> 

Reply via email to