[aprssig] 9600? Faster?

[Tim Cailloux]

Earl Needham wrote:

>         Well -- are we stuck at 1200?  Or should we try something even 
> faster than 9600?

I work on mobile data for a career, and 19200bps is about the limit of 
what you can do with one receiver and still get decent range in a land 
mobile environment, with a FEC optimized for mobile data.  Faster speeds 
require some sort of diversity receive.  It comes down to a decision 
between capacity and coverage, and it is always a trade-off.  9600bps 
buys you more users than 1200bps, but the radios have to be set up to do 
it to gain any advantage.

I think someone on the list posted that the D700s run a 250ms TX on 
delay (or some large number N, where N >>> time to transmit the data at 
1200bps).  Why run 9600bps when the speed increase on a per-transmission 
basis is negligible compared to the overall packet transmission time?  
Running faster data speeds will require some optimization of the 
radio-modem interface and careful selection of the radios used in the 
environment.  Even among the land mobile voice radios, the fastest a lot 
of radios could reasonably go is less than 9600bps.  If amateur 
transceivers are designed around voice and 1200/2400bps packet, I 
wouldn't say it's reasonable to expect all that many to do 9600bps out 
of the box, regardless of the presence of a 9600bps pin on the accessory 
jack of the radio.

On a smaller scale, I certainly think that users can build a 9600bps or 
faster setup (and I enjoy seeing it done), but large scale the chances 
of higher speed seem limited.

(I only wish I could retrofit my company's 43.2Kbps stuff for APRS and 
amateur data!)

tim

-- 
Tim Cailloux
tim at cailloux.com