Re: Digital privacy

> >Pssst, what about privacy? Again, since your voice is digitally 
> >processed, it's >less likely that anyone could eavesdrop on 
> >your conversations. 
>  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 
>  
> ?Como?  This is absolutely amazing! 

Well, it is true, from a pragmatic point of view.  A neighborhood kid with a 
Radio Shack scanner can't listen in on digital cellular calls.  You have to 
actually hack a phone, which is a much less prevalent skill.

I still think that CDMA+DES is the way to go for secure cellular, but from a 
purely pragmatic point of view simply going digital does increase privacy.  
Using analog cellular is like using a walkie-talkie.


Amanda Walker
InterCon Systems Corporation

I wrote up a short "test bed" for Dining Cryptographers over IRC (Internet
Relay Chat).  The implementation is not secure, as it uses PRIVMSGs to
exchange flips between neighbors, but some good soul with a bit of time
could go in and add DES encryption to them. 

The client allows for anonymous roundtable discussion between clients on 
a single IRC channel.  The key-sharing graph is a ring, so it only takes 
two to collude to find what a particular person is sending, but that can 
be expanded also.  There is no collision detection, but it is usually 
pretty obvious (i.e. if you send something and don't get it back, you 
know a collision happened).

I also included a paper which has some discussion and references to 
making the DC implementation better, including things like protection 
against attacks coming from the IRC server itself, protection against 
someone sending stuff all the time disallowing service, etc.

This has all been put in the cypherpunks ftp incoming directory.

-Thomas

> 
> > >processed, it's >less likely that anyone could eavesdrop on 
> > >your conversations. 
> >  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 
> >  
> > ?Como?  This is absolutely amazing! 

 Amanda Walker writes:
> 
> Well, it is true, from a pragmatic point of view.  A neighborhood kid with a 
> Radio Shack scanner can't listen in on digital cellular calls.  You have to 
> actually hack a phone, which is a much less prevalent skill.

	Both TDMA and CDMA cellular use digital echo surpressors which
means that a simple hacked phone will only recover the base (landline)
end of the conversation - almost all traces of the mobile end of the
conversation will be eliminated by these DSP devices which are required
to eliminate the otherwise very noticable echo due to the O(100 ms)
delays of the vocoders used. 

	Recovering the mobile end of a nearby call which is usually easy
with FM analog cellular (AMPS and NAMPS) and a scanner, and usually
unnecessary anyway due to the low return loss (high echo) of the
wireline trunks and switches which makes the mobile caller's voice
clearly audible on the base station transmission, will not be possible
for either CDMA or TDMA using a hacked phone as both systems use more or
less entirely different modulations and transmission techniques in the
mobile to base direction.   These modulations cannot be recovered by
simply hacking the firmware of a phone - they require different signal
processing electronics.

	And CDMA uses strict power control in the mobile to base direction
which ensures that much of the time the mobile signal will be well below
the threshold of detectability at a listening post located anywhere but
very near the caller.

> 
> I still think that CDMA+DES is the way to go for secure cellular, but from a 
> purely pragmatic point of view simply going digital does increase privacy.  

	It greatly increases privacy against casual snooping, but of
course does very little to protect against the kind of serious threat
that both the TLA's (ours and theirs) and large criminal and industrial
spy operations pose.  As such it may lead people to be more careless
because they have never seen the risk of cell call interception demonstrated
and delude themselves into believing it is not possible.  I hope that
what interception is possible with hacked phones becomes widely visible
so the illusion of security is not regained when the digital switch happens.

	And of course I repeat old news when I point out that the NSA and
other TLA's have been quietly fighting a battle for years to keep cellphones
from using effective encryption, and have so far blocked it out of the
standards.

> Using analog cellular is like using a walkie-talkie.

	Against serious threats, using any clear phone is like using
a walkie talkie.   Against nosey neighbors using an analog cellphone is of
course not advisable for any kind of private conversation.  It is certainly
true that most people are much more threatened by their nosey neighbors than
serious spies, however, and so digital cellphones will make calls much more
private for ordinary people.

						Dave Emery  N1PRE
						die@die.com

> Well, it is true, from a pragmatic point of view.  A neighborhood kid with a 
> Radio Shack scanner can't listen in on digital cellular calls.  You have to 
> actually hack a phone, which is a much less prevalent skill.

   True.  And the fact is, monitoring any _one specific call_ is rather 
difficult to do without specialized equipment - in my experience with 
modified analog phones, one has to hop between 50 cells at the minimum 
before they can hope to monitor a specific call.  I've yet to obtain a 
digital cellphone to modify, but with the recent legislation about 
cellular phone monitoring, I wouldn't be surprised if it were more 
difficult to hack them.  (and just from the size of them alone, I would 
guess that they're mostly surface-mount, which is a pain in the arse for 
most people without elaborate facilities)

> I still think that CDMA+DES is the way to go for secure cellular, but from a 
> purely pragmatic point of view simply going digital does increase privacy.  
> Using analog cellular is like using a walkie-talkie.

   Good analogy. 

-jon
( --------[ Jonathan D. Cooper ]--------[ entropy@intnet.net ]-------- )
( PGP 2.6.2 keyprint: 31 50 8F 82 B9 79 ED C4  5B 12 A0 35 E0 9B C0 01 )