Further updates to this package should be available from CPAN:
http://search.cpan.org/AUTHOR/A/ALEX/TDMA/
#
# This module is not intended to require 64-bit support
# but the objects it creates can rapidly exhaust your
# given slice. This means you may run out of ram and your
# test may fail, despite the API still working. Right now,
# there is a hack in place to prevent failure during
# automated testing.
#
# Nothing other than my own bugs should prevent you from
# using this code the way it is, if you want to build
# partial epochs or see how the TDMA stack is composed
# (I had to reverse-engineer a stack of *printed* powerpoint
# documents!). So, this is pre-release software, and thar
# be dragons. I'm fixing it, but it will be slow unless
# somebody would like to either help or get this more complete.
# Right now it needs a lot more tests.
#
# Alex J. Avriette, April Fools' Day, 2010.
#
The purpose of this project at the moment is to teach more than to be
used for communications. On the computer I initially began development
on (A 2ghz Macbook), it was nowhere near fast enough to actually work on
an epoch and be useful. I now develop on a Macbook Air, which is
substantially slower. So, if you find a way to marshal data and build
your objects fast enough, you can in fact speak TDMA to whatever you'd
like.
TDMA is pretty cool in that, provided you have key escrow, it's damn
difficult to break as encryption (because of jitter, propagation, and
other tricks I am sure people will come up with – like jitter and
propagation that look an awful lot like HTTP traffic or something).
Even without transmission start keys (or "crypto ignition keys,"
"certificates," even ssh identities, whatever you want to call them --
you must have them before you start talking), the communication happens
across either one channel with many, many slices, or it happens across
many, many channels with relatively few slices per each channel. This
means that anyone attempting to listen in has to know how to reassemble
the traffic – by knowing the epoch or by having sufficient data to make
a "guess" – and do so in a fashion that is both as quick as yours
(suppose you're transmitting at 3gbit and they're only able to pull it
down at 100mbit), and is smart enough to do the reassembly in real-time.
If you are willing to accept a little bit of loss in your communication,
you can even arbitrarily drop packets – or insert "junk" packets which
you don't want anyways – to reduce the amount of time you spend actually
processing the data (as mentioned elsewhere, this is processor
intensive, depending on your data).
So, really, this is a framework upon which you can build very fancy,
very hard to break communications protocols, or you can just play with
the code and learn how TDMA works.
TDMA, for what it's worth, has been out of use for many years in cell-
phone usage, so you're not likely to get very far there, and you're
probably not going to get a chance to speak to an F-16. You might be
able to listen, but again, without knowing the epoch, having the keys,
and getting every single packet, it's really tough.
I'd also like to say that I did not use any classified military data in
building this software. I did not build it with the intent to disrupt
traffic between military vehicles or between cellular providers and
their customers. I built this module to teach myself how TDMA works,
because I just learn better through programming. I really hope you don't
go and use this software to mess with the military. They don't like
that.
This software is best used as a teaching tool than as a piece of software
as such. To get the speed you would need, you would require FPGA-like
speeds, and perl, even on sixteen-way-Macintoshes, isn't going to be
fast enough. Sorry.
It might teach you how to do it with multiplicity in C, and if that's
the case, it might be fast enough and you might just come up with the
new OTP uncrackable point-to-point system. Good luck with that, but
don't let me know. I don't need the fibbies asking me why I felt it a
good idea to codify TDMA in perl.