Andrew Gideon wrote:
It's somewhat like Ethernet. Listen before talk, but you still can have
collisions. Collision detection is problematical (listen after talk).

If I recall correctly, Ethernet uses "listen while talking". If what is
sent isn't what is heard, collision is presumed and the sender backs off.
There's also a problem, again: if I recall correctly, in the back-off. I
think that the problem was potential starvation as a sender could be
repeatedly forced to back-off.

That's correct. "Listen while talking" doesn't work for ADS-B. In
fact, the only part that is really of value is the "Listen before
talking". The backoff problem you refer to is the "knee" in Ethernet
performance. One big advantage of Ethernet (over the other types) is
that the performance for a lightly loaded line is GREAT. It then tends
to drop off just a little bit worse than would be forced simply by the
amount of time each additional sender requires... until you hit a
"knee" where the performance starts to degrade very very badly with
each new load placed on the line. Basically, as you state, almost
every packet sent gets a collision with someone.

Fortunately, ADS-B does have the "burn through" phenomena - something
that wired Ethernet doesn't have.

It's why one should select non-overlapping channels for neighboring clouds.

I wish the FAA would show the same sense. But they are bound and
determined to fit the entire ADS-B thing into MODE-S, in spite of
existing MODE-S congestion in many areas already. And in spite of the
known fact that MODE-S can't handle the MANY other potential benefits
of ADS-B, like weather uplink, etc. I hope they don't force it, but
having forced MODE-S onto the airspace system over great opposition,
they are politically pushed to find a use for it. [TCAS doesn't count,
since potentially ADS-B surplants much of that function, especially
when full IN/OUT modes are implemented.]

But what about the ground stations' transmissions (ie. the TIS-B and FIS-B
streams)? Are the ground stations' transmitters sufficiently powerful so
as to "punch through" aircraft-aircraft traffic? Is there some other way
to give the ground stations' broadcast at least a decent chance of being
received?

Hey, this is a government program. You want perfection? G

Seriously, this is why some suggested that the real solution was to
bypass the transmitting part of the ground stations. Instead, take the
estimated $1 Billion savings from eliminating the present VOR system
(except for a small backup infrastructure) and just BUY every plane out
there an ADS-B system.

No, I'd thought that this was all already defined. How else could these
services be running in Alaska and part of the US East Coast?

You'd think so, wouldn't you. But no, it's not already defined. Even
the frequency (MODE-S vs. VHF) are still up for grabs. All those
systems currently running are "test systems" only.

BTW, the latest edition of AVIONICS Magazine had some pretty good
articles in it on ADS-B.