[aprssig] The future of APRS. (Was Re: weak signal ISS packet)

[Dave B]

Morning.

Comments inline.

From:           	Steve Dimse <steve at dimse.com>

> On Jan 13, 2015, at 4:43 AM, Dave B <g8kbvdave at googlemail.com> wrote:
> 
> > why the heck are we 
> > not moving Packet/APRS forward to use faster but robust modulation
> > schemes and methods?
> 
> I can think of three reasons offhand. First, though far from perfect
> and farther from state-of-the-art, the system works for most people
> most of the time.

Most is not all.  (But nothing is 100%)


> Second, it is far easier to create a new system and attract users than
> to entice people to switch to an incompatible system. That's why
> Windows is still around.

Irrelevant.  We're not comparing OS's.

Windows is still about, because it's pre-installed on just about every desktop and 
laptop PC you buy from the mainstream retailers, and the business support (IT) 
trade is entrenched in it, it's a cash cow to put it bluntly.


> The third you gave yourself: "I wish I knew how to do that sort of
> stuff myself." Implicit in this answer is "and I'm not going to learn
> how". Maybe you aren't smart enough to learn how, but I doubt that.
> Instead, you have things that are more important or that interest you
> more, and you choose not to learn how. 

I still work for a living (I can't afford not to) and have a family life.   There are 
not enough of,  "the right sort of contiguious hours"  in a day as it is.

I can learn (and still do!)   But it takes time, and though I know some of the 
basics of DSP, I know pitifully little about how to code the stuff.   (For whatever 
reason, I have a mental block regarding threaded programming, and "streams".)

Maybe when I retire, if the Government ever let me retire that is.


> No one makes money on APRS development. We do things like put ads on
> our web sites and charge for software in the hope of stopping the
> hemorrhage of money we spend to develop and provide our service, maybe
> even earning a few cents an hour on our time, or providing some
> justification to significant others for the hours away from them. But
> any way you look at it, this is something you do as a hobby because it
> interests you. So, if it isn't something you are willing to waste
> thousands of hours (and dollars) on, and that decision is repeated a
> few ten-thousand times by everyone else in APRS, it doesn't get done.

That hasnt' stopped things developing in other areas of Ham Radio, in particular 
the proliferation of SDR software.  Many digimodes, some of which could 
potentially knock spots off 1200bd AFSK (RPR* etc)   Digital Voice (FreeDV etc) 
and digital SSTV etc, most developed by groups of unpaid people, though with a 
core few who do the coding, but many who do the testing.

(* RPR is sort of semi propriatary, but tantalisingly some information about it, is 
in the public domain.)

I don't know for certain, but I think G3RUH developed the "RUH" 9600bd 
modem, much in his own time with others who I forget, that was poorly adopted 
(except in small areas) it is however, included as standard in many Kenwood 
APRS capable radios.

APRS (sadly) is stuck in a backwater in many people's eyes, compared to what 
can be done now, "at very low cost" with a mobile phone, that just about 
everyone in the developed world now carries with them nearly all the time.

The best single thing that has happened in recent times to keep APRS going 
outside of the US for most people (and I suspect that Bob B may not agree) is 
the aprs.fi website.


Meanwhile, for the most part, we're still stuck with an antique AFSK system 
designed for voice phone lines, where as has been recently highlighted, there are 
major problems with the way it's used with "normal voice" radios to handle the 
link.

Not to say, when "correctly" setup and applied it doesn't work well, it can and 
does, but there are so much better alternatives now available.    Keep the AX25 
protocol, that is time tested, it's the on-air modulation schemes that could do 
with seriously updating.


What I was trying to get across, is that as most amateur and ex PMR voice radios 
have largely the same pre/de-emphasis characterstics, if we are going to keep 
1200bd AFSK, lets sort out the modems (software and/or hardware) to cater for 
the amplitude/phase distortion that causes.   Maybe, it can be done with some 
deftly designed analogue allpass filter before the tones are sent to a radio?

You won't change the way the likes of Kenwood do it, feeding the tones directly 
into the modulator, and taking the audio from the discriminator, bypassing the 
voice pre/de-emphasis "features", but perhaps it could be possible to make 
what's on air much much better.


Similar to this.    "Back in the day" of CUTS format digital data on cassette tapes 
for "home computing".   A simple allpass filter came about (by who I don't know) 
that when applied between the cassete player's output, and the computers autio 
input, dramatically improoved the reliability of the decoded data, to the extent 
that for the same tones, it was possible to double the data rate, and keep it 
reliable.

Many people at the time had "Why didn't I think of that" moments, as when you 
understood what the problem was, the solution was obvious.  (One OP-AMP, and 
a handful of passives.)


> Steve K4HG

	73 Steve.

(And yes, I do know who you are, and what you have done in the past for the 
Amateur community as a whole.   Much appreciated by many, me included!)

Dave G0WBX.