[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[HLL] Re: VP1 Comparison

Status: RO
Content-Length: 2893
Lines: 70



> > We would like to use a standard-144-modem.
> > These modems use V32bis and are therefore synchron "on the line".
> > However VP1 has a direct "synchron" connection to the modem-chip.
> > In our case this synchronous data is converted to asynchron.
> > While this does not change the baud-rate (we can still use gsm-full-rate),
> > it changes several aspects for frame-synchronisation.
> > These differences to aquire and maintain sync were what i meant.

> You will find that said differences will push your protocol overhead just
> high enough to no longer being able to use the GSM full rate codec.

Uops - i really don't hope so !
This is exactly what we want to do...

> But I agree with you that anything that will run on just a standard
> Pentium with a standard modem will have to use an async protocol.

...because we want to be compatible with pentium/modem/async144 !

> Since I apparently missed the beginning of the discussion, would you
> please explain to me where the device your are building gets into play? If
> you run the codec and the crypto on a regular PC, what do you need the box
> for? Or is the box only used for stand alone mode when you don't have a PC?

Yes !

Our box will be a stand-alone-secure-phone.
It contains a 32-bit risc-cpu (30MHz) to run crypto, gsm and local stuff.
It also contains some kind of analog handset, either directly or by
emulating standard phone line and using standard-phone as analog in/out.
However it does _not_ contain modem-chips - instead we provide rs232-if.
You plug in a standard 144-modem. This avoids legal problems and simplifies
design. The box can also contain a csprng in hardware and chip-card-interface.

Those strange people from holland ;-) work on pc-based software to build
a compatible device using pentium/soundcard (or whatever...).
We want to share a common data-fromat for those "async" modem-links.
But this pc-based implementation can also make use of tcp-links.
The stand-alone-device will not support this mode initially.

! So one important thing now is to define the "async" modem data-format !

As part of this job i had a look at the specs for "VP1".
It looks as if we can not support their "modem-standard".

But there is a "quick and dirty" proposal for a async-low-level-format,
taking into account async nature, 144-bandwidth and gsm-requirements.

You probably missed it. I send it to you via email.

Comments/enhancements on/of the subject/proposal would be nice !


Version: 2.6.3i
Charset: latin1