Table of Contents
IP and Packet Radio
It has been possible for a long time to run IP over packet radio.
There are some limitations:
- At anything but high speeds (e.g. 4800bps and up), modern TCP/IP implementations perform poorly since the higher layers in modern operating systems have now typically been tuned for high bandwidth, low latency, low loss network devices. This results in a poor user experience.
- In many countries, including the UK, encryption is not permitted on the amateur service, so no SSH (unless running with null ciphers), no HTTPS, and so on
- This is likely to work spectacularly badly where there isn't a solid underlying AX.25 link. Tune the link for zero retries first, before moving up the stack to IP.
But it does fundamentally work, and might have niche applications, and it's also a bit of fun to be had.
There are at least two different ways to do IP over packet radio:
- Linux kernel AX.25 support +
kissattach tncattach
and I've written them up briefly below.
If you know of other ways, please feel free to extend this page.
Methods
Linux kernel AX.25 support + kissattach
Note that over time, this has become a bit broken, but the good news is there are recent fixes still making their way into distributed Linux kernels.
That said, on a modern Debian-ish Linux:
sudo apt install ax25-tools
Define /etc/ax25/axports like so:
# /etc/ax25/axports # # The format of this file is: # # name callsign speed paclen window description # myport M0LTE-1 57600 255 7 NinoTNC
In that second column, set a unique callsign-ssid at each end, e.g. M0LTE-1 and M0LTE-2. 57600 is the correct serial port baud rate for the NinoTNC, adjust as required. (Any KISS modem can be used.)
Then, on each system:
sudo kissattach /dev/ttyACM0 myport 10.0.0.1
Specify a unique IP in the same subnet at each end, e.g. 10.0.0.1 and 10.0.0.2. You can use private (RFC1918) addresses, or any others you have allocated, eg from 44net.
/dev/ttyACM0 is the device path for a single NinoTNC, adjust as required.
After that, the interface ax0 should exist:
tf@testnode1:~ $ ifconfig ax0
ax0: flags=67<UP,BROADCAST,RUNNING> mtu 255
inet 10.0.0.1 netmask 255.0.0.0 broadcast 10.255.255.255
ax25 M0LTE-1 txqueuelen 10 (AMPR AX.25)
RX packets 102 bytes 10782 (10.5 KiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 153 bytes 30136 (29.4 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
and you should be able to ping down the tunnel:
tf@testnode1:~ $ ping 10.0.0.2 PING 10.0.0.2 (10.0.0.2) 56(84) bytes of data. 64 bytes from 10.0.0.2: icmp_seq=1 ttl=64 time=345 ms 64 bytes from 10.0.0.2: icmp_seq=2 ttl=64 time=344 ms 64 bytes from 10.0.0.2: icmp_seq=3 ttl=64 time=342 ms 64 bytes from 10.0.0.2: icmp_seq=4 ttl=64 time=349 ms 64 bytes from 10.0.0.2: icmp_seq=5 ttl=64 time=347 ms ^C --- 10.0.0.2 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4006ms rtt min/avg/max/mdev = 342.304/345.252/348.568/2.272 ms
and, for example, serve up web pages:
tf@testnode2:~ $ sudo apt install nginx -y
tf@testnode1:~ $ curl http://10.0.0.2
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
<style>
body {
width: 35em;
margin: 0 auto;
font-family: Tahoma, Verdana, Arial, sans-serif;
}
</style>
</head>
<body>
<h1>Welcome to nginx!</h1>
<p>If you see this page, the nginx web server is successfully installed and
working. Further configuration is required.</p>
<p>For online documentation and support please refer to
<a href="http://nginx.org/">nginx.org</a>.<br/>
Commercial support is available at
<a href="http://nginx.com/">nginx.com</a>.</p>
<p><em>Thank you for using nginx.</em></p>
</body>
</html>
There is more to the Linux implementation than this, for example setting port parameters to tune the link with kissparms, setting this up as a systemd unit to start at boot, and so on.
It may be worth denying dhcpcd access to ax0:
# Add the following statement somewhere at the beginning # of /etc/dhcpcd.conf to prevent dhcpcd from changing MTU denyinterfaces ax0
To bring down the interface:
sudo pkill kissattach
tncattach
A program exists which is independent from the Linux kernel AX.25 support / kissattach, called tncattach.
tncattach is available from https://github.com/markqvist/tncattach
Once compiled, this is a one-liner at each end to set up:
sudo tncattach -i 10.0.0.1/24 -m 500 -n -s M0LTE-1 -t 600 -v /dev/ttyACM0 57600
and
sudo tncattach -i 10.0.0.2/24 -m 500 -n -s M0LTE-2 -t 600 -v /dev/ttyACM0 57600
This will set up a tnc0 network interface.
To compile tncattach, on a Debian-ish OS, follow the steps in the README, reproduced here for convenience:
sudo apt install build-essential git clone https://github.com/markqvist/tncattach.git cd tncattach make sudo make install
There's some really useful information in the README around setting up point to point links, blocking IPv6, setting MTU, station ID and so on.
It may be worth denying dhcpcd access to tnc0:
# Add the following statement somewhere at the beginning # of /etc/dhcpcd.conf to prevent dhcpcd from changing MTU denyinterfaces tnc0
Note that as far as I'm aware, this does NOT interoperate with stations running IP over the Linux kernel AX.25 stack.
Link testing
This will give the link a reasonable workout:
tf@testnode1:~ $ sudo apt install fping -y
tf@testnode1:~ $ fping 10.0.0.2 -M -b 200 -s -o -l -Q 10
[12:38:16]
10.0.0.2 : xmt/rcv/%loss = 10/10/0%, outage(ms) = 0, min/avg/max = 584/642/896
[12:38:26]
10.0.0.2 : xmt/rcv/%loss = 10/10/0%, outage(ms) = 0, min/avg/max = 612/614/619
[12:38:36]
10.0.0.2 : xmt/rcv/%loss = 10/10/0%, outage(ms) = 0, min/avg/max = 614/614/615
^C10.0.0.2 : xmt/rcv/%loss = 35/35/0%, outage(ms) = 0, min/avg/max = 584/622/896
1 targets
1 alive
0 unreachable
0 unknown addresses
0 timeouts (waiting for response)
36 ICMP Echos sent
35 ICMP Echo Replies received
0 other ICMP received
584 ms (min round trip time)
622 ms (avg round trip time)
896 ms (max round trip time)
35.000 sec (elapsed real time)