[aprssig] most packets are transmitted blind

[Matti Aarnio]

On Wed, Sep 23, 2009 at 05:41:04PM -0400, Robert Bruninga wrote:
> >> there are 350 other user packets flying 
> >> around and most of them are blind transmitters 
> >> that could care less what they collide with.
> > 
> > Do you have data to support this?  

...
 
> highest APRS densities on the planet).
> 
> Now then lets say it is PRIME MOBILE time, and all 5 other
> mobiles I can hear direct are on the road, and beaconing at a 2
> minute rate.  My station (listens first), and AVOIDS those 5 out
> of 120 time slots.  But when I hear silence, I have an equal
> probability of colliding with any of the other 241 stations.
> And there is nothing I can do to prevent that.

There is much.  Have dense RECEIVER network feeding far fewer
digipeater TRANSMITTERS.  We Finns are building such things.
(How do we connect receivers to transmitters?  We use Internet.
If the internet fails, the receiver at digipeater transmitter
site is only one to be used, but so be it...)

With many receivers at diverse locations, _some_ of them will read
packet from party A, others from party B, etc., and central transmitter
can then decide to transmit them, or drop as duplicates.

> People need to understand how the ALOHA Packet channel with
> digipeaters work, not just take the common platitudes and try to
> apply them where they don't apply.

If you insist on using single digipeater RTX at a bell-tower, or
other similar simple schemes, then sure.

On non-trivial terrain/obstruction conditions, the Raleight Fading
Universe is quite complicated.  Furthermore, adding more transmitter
power to your car tracker just changes things to your favour in
such a way that soon others too need to have 50 W or 100 W for
tracker transmitter, instead of doing well with 5 W.

The Digipeater can use 30-50 Watts output.
The Digipeater is an "Aloha Master" in a sense.

It is all about "cell size", where each transmitter signal gets
below other signal at the receiver by that magic "FM treshold"
(8 to 10 dB stronger signal masks away all other signals)

However frequency errors in ham radio transmitters dominate over
the "FM treshold", and moving transmitters received signal pumps
10 to 30 dB due to multipath interference, and other similar reasons.

AFSK demodulators are also rather worrying piece of kit..  Some
need excellent noise free signal, others manage with rather weak
S/N.   Lets say just that telephony modems are not always the best
for packet radio, but "DSP" modems are not coming on small solderable
chips...

> This topic was not about trackers at all.  It was simply a
> statement of fact about how the APRS network works and how it is
> a collision limited system, and so packets should be kept as
> short as possible to maximize one's OWN probability of getting a
> message packet through.
> 
> I was saying it was ill-conceived for people to be proposing 256
> byte English messages and therefore 1024 byte long foreign
> language packets on the national APRS channel where BREVITY is
> the key to success!

You did that yourself - 63 CHARACTER MESSAGE of very high code-
point value all of them --> 256 byte message.

Lets stick on original:  Message size limit is 67 (something) BYTES.
Chinese write there 20 WORDS.

> That's why we call it text-messaging...
>
> Bob, Wb4APR

Matti, OH2MQK