End-to-end encrypting US GSM phones?

[Steve Schear]

At 11:59 AM 12/31/2001 -0500, Trei, Peter wrote:
>Ryan suggested earlier that the Ipaq with Fireball isn't up to the task.
>I don't think this is correct. I point you to
>
>http://www.stanford.edu/class/cs344/staff/group10/final_report.pdf
>
>in which some Stanford students produce a wireless ipaq2ipaq
>VoIP system. Biggest problem is that it's simplex, not because
>of cpu limitations but rather that the Compaq HW can listen or
>make noise, but not both simultaneously. Using a headset which
>interfaces through the expansion port might enable you to
>avoid this problem (and the BT headset could also offload the
>codec). A 206 MHz Ipaq would, I think have sufficient HP to
>do the crypto.

Agreed.  The port expander approach should be explored.  Also needed is 
WinModem control for smooth analog (clear) to digital (crypto) transition.


>I've toyed with the idea of porting SpeakFreely to Ipaq, but have
>not actually done anything at this point.

Several developers, including an ex-Starium engineer, and I have been 
talking about just such venture.  I sure between all of us we have the 
wherewithal to make this happen.

Besides the half-duplex problem some changes to the SF protocol are desired 
to work better with SOHO firewalls now common.

steve