I removed all the unicode from the landing gophermap because I realized that it wasn't going to render correctly on older hardware. I also tried to stay under 40 lines wide where possible.
Meanwhile I'm trying to play with gr-iridium still and L-band satcom stuff.
While you can just run the scripts in iridium-toolkit and get some degree of success,
I found that tuning the params can make quite a difference.
A few observations:
* Keep an eye on the OK rate as well as the OK percentage. Obviously if you're receiving lots of stuff
there's going to be lots of weak packets that don't decode 100% - this is not necesarily bad.
That's where you want to keep an eye on your rate of OK frames/sec. I wrote a little script that
dumps this in a terminal based graph in real-time, looks like this:
_____ ___ _ __
|___ / _ (_)/ /
/ / | | |/ /
/ /| |_| / /_
/_/ \___/_/(_)
_ ___ __
/ |/ _ \ / /__
| | | | |/ / __|
| | |_| / /\__ \
|_|\___/_/ |___/
^
18 +---------------------------------------------------------------------+
| AA + + + + + |
16 |-+ AAA "-" A +-|
| AAAA AA AA |
14 |-+AAA AAAAAAAAAA +-|
| |
| AAA AA AAAAA |
12 |-+A AAAAAA +-|
| A AAAAAA |
10 |-AA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA +-|
| A |
8 |AA +-|
|AA |
6 |AA +-|
|A |
| |
4 |-+ +-|
| |
2 |-+ +-|
| + + + + + |
0 +---------------------------------------------------------------------+
0 2000 4000 6000 8000 10000 12000
I'll dump this script in the radio/iridium folder on this gopher server (it's under projects).
(it's called rate_ok.sh)
* Obviously your LNA gain is a huge factor, combined with the 'threshold' setting which
defaults to 8.5 dB over noise, but might need tuning, especially if you muck with the LNA gain.
* Even though the documentation claims there is only marginal benifit out of decimation < 4, I
am getting much better decoding success and rate out of decimation=2 (1 over 2 has little effect)
* I am getting bad results if I use anything other than the default sample rate supplied in the rtl
example config, even though rtl dongles should be able to handle a higher sample rate.
Since the sample rate is used in a lot of the timing math - it might be that it throws something off or
that something else needs to be tuned along with it.
* Looking through the code, it turns out that the only difference between 'offline' and 'online' mode is
that online mode may drop frames - i'm actually getting better results in the voice decoder if I
run iridium-extractor with --offline
* --multiframe (turn on multiple frame per burst support) sound good and seems to have no averse effect.
In other news, I'm also still trying to think of ideas on how to get something out of inmarsat on linux.
There isn't much documentation out there and all the tools that do exist are for Windows and closed-source.
Oh, also, the kind people at floodgap were kind enough to list my gopher server. \o/