[aprssig] alt input digipeater implementation (brain storm)

[Wes Johnston]

Just to get people thinking... here are a 7 ways to implement an 
alternate input frequency digipeater.  It can be extremely simple or 
very tightly coordinated.  In the end, it is up to the owner of the 
digipeaters to weigh the pros and cons of each method.  This gets to be 
like cars.... Speed (network reliability) is money, how fast do you want 
to spend? 

The whole point to the alt input digi is to give the locals a clear shot 
into the local digipeater... a priority "back door" into their local 
digi with less congestion and less of a chance of colliding with another 
local.

In any case, try to use a digipeater that is capable of callsign 
substitution so you can easily identity the entry point in the network.

We'll start with the cheap-charlie method.  This will work areas that 
have little locally generated traffic, and lots of DX coming in.

1)Take a digi in your area with reasonable coverage, and simply program 
the radio to listen on 144.99 with a -600 offset.  This method will 
stomp over any traffic on 144.39 when it transmits, but if the users in 
the area decide that they can live with the loss of the occasional DX 
packet, ok.

2)Take a digi in your area and program it as above 144.99/-600.  Place a 
2nd reciever on site that can hear 144.39 and use it's squlech 
indication hooked to the XCD pin on the digi's TNC.  Disadvantage: your 
TNC can't hear any packets on 144.99 that come in while it's digipeating 
a packet out on 144.39.

3)Hook a TNC up to a receiver on 144.99, and a transceiver on 144.39.  
Create a Y cable for the TNC to allow it to connect to two radios... one 
to listen, one to TX.  Use the 144.39 radio squelch indication hooked to 
the tnc's XCD pin.  Use 2m 600khz cavities just like voice repeaters 
do.  Disadvantage: duplexer cavities are expensive.  Advantage: the TNC 
can contine to receive packets while it's digiepeating them on 144.39. 

4)Use 2m 600khz duplexer cavities to allow simultaneous operation on 
144.39 and 144.99.  Use a multiport digi program such as digined or 
ui-view to collect data from a TNC on 144.99, and also provide full 
serivce digi on 144.39.  Disadvantage: duplexers are expensive.  
Advantage: everything is at one site.

5)Create a cross band digi.  TNC listens on 144.99, transmits on 
440mhz.  Add a 440mhz receiver to your local 144.39 digipeater site 
(which is a few miles away).  Use multiport digipeating program such as 
digined at the main digipeater site.  Disadvantage: must make changes to 
the local wide digi.

6)The cross band digi can be extended to other satellite recieving sites 
on 144.99 all digipeating data on 440mhz uplink to the local digi.  If 
all 440mhz stations can hear each other, they can avoid keying over one 
another.  Advantage to this system is that if two cars on opposite sides 
of town TX on 144.99 at the same time, both packets get delivered to the 
digi via 440mhz.  Since the two alt input sites can hear each other on 
440mhz, the first one to digipeat will cause the 2nd to hold off. This 
is expandable until the remote stations stop being able to hear each other.

7) If we extend idea #6 to the point that all stations on 440mhz can't 
hear each other (let's say they are so far from the main digi that they 
have to use a beam to get to it, and thus can't hear other stations off 
to the side of their beam), then we need to implement some sort of 
slotting system.  This could be down by using a synched clock source 
such as GPS, or clock derived from a local "good time" FM radio station, 
and using a (ahem) basic stamp to assert XCD on the TNC and release it 
when that TNC's coordinated "slot time" comes around.  For example, if 
you have 5 remote reciever sites, you'd program each one to transmit 
every 5th second (assuming packets are one second long, tops).  A DAMA 
variantion would also work here... the central digi could poll all the 
remote stations for traffic instead of allowing them to depend on clock 
synch.

Wes