About GSM/Wifi handover
wally.ritchie at gmail.com
Wed Jul 25 21:13:13 CEST 2007
On 7/25/07, Sébastien Lorquet <squalyl at gmail.com> wrote:
> Hello again,
> I'm continuing investigations about wifi handover.
> It's called UMA and is documented in 3GPP TS 44.318 and 3GPP TS 43.318
> To say it fast, the goal is to route mobile calls using an IP/IPSec link to
> a gateway server.
> I currently have 2 blocking problems.
> The first is the ability to get authentication credentials from the SIM. The
> EAP/SIM protocol documented in previous specs explicitely asks access to the
> SIM card algorithms A3/A8, but the only available interface to neo's modem
> is the AT line.
> I had a discussion minutes ago on IRC with XorA; according to him, it's
> possible if we can run ISO7816 commands with the AT interface to drive the
> card itself.
> Do you know any possibility to do this on neo's modem? Are the specs
> available or not?
I will try to answer as best I know. I cannot guarantee the following
There is a defined AT+CEAP command which supports this. It is not supported on
the present chipset and it may require a 3G USIM.
Direct control of GSM class commands on SIM is not supported by the
chipset (see earlier posts from Harold on this list). I seem to remember this
being prohibited to external TE control. It would be nice to have a tunnel to
the SIM card (two way) for other reasons e.g. support for the Bluetooth SIM
profile. Perhaps this may appear in later versions of the chipset.
> Secondly if we suppose that step is solved, it will be necessary to shut
> down the modem without sending a DETACH message! Will it be possible with AT
> or not?
I am not aware of any at present. It could fit into the framework
pretty easily. There
may be a vendor specific proprietary command to do it. As you probably know from
reading the specs there are a lot of issues in the handover from OTA
to public IP.
> Please consider this as a feasability study only, I'm currently
> investigating ways to get UMA on an open platform. Don't speak too much
> about carriers implication in the subject now, this is not yet an issue
> since I want to know if it's technically possible before going further.
Tunneling to the MSC is coming. Its in everyone's interest.
> Sincerely yours, (and sorry for my bad english, not so easy to describe
> technical things without translating french thoughts ;) )
I know - you have too many words in French ;)
> Sébastien LORQUET.
More information about the gsmd-devel