I thought I would put down on paper my thoughts about the current FDM
implementation. No performance charts or anything like that, just a binary
bit placement.
The QPSK is of course fed 2-bits, which are used to create one of four
phases to one of the sixteen subbands (carriers if you like). You might ask
yourself, are all subbands equal, and the answer from an interference
perspective is no. We've all had people creep-up to our signals and get too
close, and the edges of our signal then get jammed.
In jamming you often analyze a protocol to find it's weakest point, and
then direct the jamming power towards that part. You might think the BPSK
pilot would be a good choice, but it has the most power. A good jammer uses
the least amount of power, so that it can hide it's activity and not become
a smoking hole in retaliation.
FDMDV is currently a very easy signal to jam. The voicing bits are very
important, as without them you would have to guess if each 10ms frame is
voiced or unvoiced. I suppose there is a mathematical dice-rolling
algorithm that can do this, but my hunch says to make it more robust to
keep Bubba and Jim-Bob at bay.
Here's the 1300 bps signal from the first of two frames:
Each subband is spaced 75 Hertz, or 125 Hz (wide):
V = Voicing (4-bits)
P = Pitch (7-bits)
E = Energy (5-bits)
F = FEC (12-bits)
L = LSP (36-bits)
Subband 0 =========================================== Bubba
Subband 1 ================== V1-V2
Subband 2 ================== V3-V4
Subband 3 ================== P1-P2
Subband 4 ================== P3-P4
Subband 5 ================== P5-P6
Subband 6 ================== P7-E1
Subband 7 ================== E2-E3
Subband 8 ================== E4-E5
Subband 9 -+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Pilot
Subband 10 ================== F1-F2
Subband 11 ================== F3-F4
Subband 12 ================== F5-F6
Subband 13 ================== F7-F8
Subband 14 ================== F9-F10
Subband 15 ================== F11-F12
Subband 16 ================== L1-L2
Subband 17 ================== L3-L4
Subband 18 =========================================== Jim-Bob
The second half frame is just L5-L36 bits.
You can see I might have stopped after the first two subbands, because with
just a 150 Hz low-power jammer I can destroy your communications. Nature
and the ionosphere also make great jammers, so forget about just having to
worry about Bubba and Jim-Bob sliding up to your signal with their 2 kW AM
signals.
There are probably a dozen ways to scramble the bits, but I would suggest
first moving the voicing bits closer to the center. I'm suggesting maybe
Subband 5 and 7, and Subband 11 and 13. Then to not put all your money on
one horse, we could pair the bits with FEC bits. So maybe V1-F1, F2-V2,
V3-F3, F4-V4. Move all those FEC bits in Subbands 10-15 around. Put some
in the I-Channel and some in the Q-Channel of the PSK signal.
Then maybe to mix it up a bit, but some of the LSP bits from Frame 2 into
Frame 1 and move some of the FEC, Pitch, and Energy bits from Frame 1 to
Frame 2.
73, Steve
------------------------------------------------------------------------------
Flow-based real-time traffic analytics software. Cisco certified tool.
Monitor traffic, SLAs, QoS, Medianet, WAAS etc. with NetFlow Analyzer
Customize your own dashboards, set traffic alerts and generate reports.
Network behavioral analysis & security monitoring. All-in-one tool.
http://pubads.g.doubleclick.net/gampad/clk?id=126839071&iu=/4140/ostg.clktrk
_______________________________________________
Freetel-codec2 mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/freetel-codec2
