PACKET RADIO: Hidden Transmitters

Steve Wolf, W8IZ@W8IZ

(This text from the W8IZ packet radio bulletin

board. It's formatted to fit a 80 character screen.)

Hidden transmitters are one of the primary causes of network collisions.
They are on the obnoxious list, right up there with beacons.
This is also one of the more complicated aspects of packet. Let's start
with an example:

Now Station A and B are both low-powered, poor-antenna low profile
stations. Station A can not hear B and B can not hear A. However, the BBS,
with its great location and equipment, hears both A and B.

A and B both check into the PBBS. The PBBS acknowledges A and B. When
it comes time for A and B to enter commands, they do it at the same time.
Station A begins transmitting. B does not hear A so begins transmitting. The
BBS gets two signals at the same time. CRASH!

What happens now is fairly amusing. There is a formula in the TNCs to
tell them when to retry the crashed packet. This formula adds a random time
delay, a little bit of extra time, to the wait for a retransmission. A and B
should have different times added to their waits. They should not crash
again. But ... they will. They will begin transmitting at times that are a
bit farther apart. Not hearing each other, CRASH!

This time, the PBBS probably decoded the header of one or the other and
sends a "no I didn't get it", a NACK, to one of the stations. From this point
the stations will all communicate ... at least until the next crash.



The higher profile a station is, the more likely the station will get
blitzed by a hidden transmitter problem. This is why the problem is most
prevalent on 145.01 and 223.7. There are lots of high profile nodes, PBBSs
and users.

A digipeated packet will add to the problem. It has the highest
priority. It does not wait. As soon as the frequency is clear, it goes. If
someone is using A or B as a digipeater, the problem will be worse. This is
the reason that digipeated beacons are discouraged. A digipeated beacon will
take precedence over all other traffic on the network.



Use a frequency away from high profile stations. This is the reason the
NCARC choose 145.09 for a users port. It is away from the high profile
stations on the other frequencies. It is also the reason that using a 220
node into the board is efficient. The high-profile node avoids the hidden
transmitter problem. The two most used routes avoid the hidden transmitter

Don't digi. Yeah, yeah, yeah. I know. But what does it take to
convince you?!? A digipeater will be faster ONLY if the network is clear of
any other traffic AND there is only ONE digipeater. Any more than one
digipeater and your path will require so many retries you will normally cause
the network to shut down.

Use the nodes that are available on almost all the frequencies. A node
can act as a traffic cop, doling out time to everyone. Everyone can use the

In Seattle, where towers come in the form of mountains, a node is so high
profile that any use of digis destroys the network. Almost everyone is a
hidden transmitter to a good part of the network. If two stations begin
digi-ing, the rest of the network leaves to watch TV.

Return to Packet FAQuette Index