🚀 go-pugleaf

In article <6596@pt.cs.cmu.edu> Andreas Nowatzyk writes:

>>There is lots and *lots* of redundancy in the human voice, and it's very
>>hard to hide it completely.

>Not necessarily true: the (west) german police uses a system that digitizes
>speech, cuts the data in short blocks, permutes the order of these blocks
>according to a digitally computed key, converts it back to analog and sends
>it.

I have heard of systems like this, and I think that some government agencies
(maybe the Secret Service) tried them for a while. The Secret Service now
uses Motorola's DES system.

Another system that is available on the open market here in the US is the
so-called "split band rolling code" scrambler. It is a combination of
digital and analog technologies, and offers reasonable voice security at
a decent price. These systems are designed to be used on common commercial
two-way FM radios.

Basically, the system spilts the voice band into two pieces (upper and
lower)....there are 32 possible split-points. When the voice band has been
split, each piece of the band is inverted, then the two inverted pieces
are recombined and transmitted over the radio channel. The neat trick is
that the split point in the voice band is changed from 4 to 60 times per
second, controlled by a "rolling code", which, from what I can gather, is
a pseudo-random generator.....the seed that you choose determines the code
sequence that splits the voice band. To maintain sync, the system transmits
a sync-burst every second or so....this allows people to join a conversation
in progress (if they have the key!). A very robust error correction system
ensures that the sync burst gets through noise.

Integrated circuit designed for the various functions needed are made by
MX-COM in Winston-Salem, NC. If you are so inclined, I feel certain that
some of the IC's available could be adapted to use in a phone scrambler.






Neal