A5 Cipher

[Owen Lewis]

> -----Original Message-----
> From: ukcrypto-admin@chiark.greenend.org.uk
> [mailto:ukcrypto-admin@chiark.greenend.org.uk]On Behalf Of Markus Kuhn
> Sent: 11 July 2002 17:42
> To: ukcrypto@chiark.greenend.org.uk
> Subject: Re: A5 Cipher
>
>
> "Owen Lewis" wrote on 2002-07-10 11:55 UTC:
> > Does anyone here have an authoritative list of the GSM A5
> cipher variants
> > (A5/0/1/2 etc) and a list of countries that use each variant?
>
> I'd be surprised if you can find an "authoritative list".

True. But you never know. I trawled from http://jya.com a nice concise if
rather old Racal Research paper describing the several GSM protocols (GSM
061088.htm)and the French and German offerings for the A5 cipher.
>
> The old GSM voice scrambling algorithms are not even mentioned in ETSI's
> "Kerckhoffs' Graveyard" list of secret standard algorithms:
>
>   http://portal.etsi.org/dvbandca/home.asp

Thanks for the URL. I don't think the GSM ciphers can ever have been SECRET.
COMMERCIAL-IN-CONFIDENCE seems far more likely. Its commercial value and
user's privacy they protect (in that order). State secrets never would have
come into it. Dear God, A5 is a *French* cipher (8-> 	I read somewhere else
that its a cutdown version of one of their old Naval hardware algorithms -
but who knows.

So, what's the vibe on the streets of Silicon Fen re. A5/3 then?

GSM's done all right. 600 million units sold and the number still rising
fast. It grabbed the future of the world's personal telecoms by the balls.
Had its cipher routines failed to protect value it would have crashed and
burned - because this would have been obvious in a manner that actual
failure to provide adequate privacy could never have been.

I read somewhere the other day that cipher for G3PP privacy is going to be
open source and a *real* cipher (and US of course). Well, we shall see if
pigs can fly. Personally, I shan't be holding my breath for it.

Owen