FTP to a SIM card? Open source IMS

Following my previous blog on the Fraunhofer FOKUS IMS event, here are some more insights from that event which you will find useful ..

Roberto Minerva of TIM Research asked the question – Why don’t we think of ftp to a SIM card? And indeed .. Why not? Applications like these are truly innovative. However, most Operators approach IMS and IP with the ‘hammer’ they are most familiar with – on the principle that “If all you have is a hammer, everything looks like a nail.” Hence we see a proliferation of voice IMS applications(where the ‘hammer’ is the voice) when we should see data

Malcolm Wardlaw(Director converged services BT 21CN ) said that many services can be implemented at a Web 2.0 layer(and strictly an IMS layer is not necessary). However, not all services can be implemented only from a Web stack. Hence, there will be always some need for telephony services – where IMS wins. This may be especially true of services that combine voice and data. Also, Identity will be a key area and a key plus point for IMS. Thus, while many services may not need the IMS stack .. But there will be some services which can’t be developed without the IMS stack and in addition, some services will command a premium by being able to provide a better class of service than the others.

The emphasis on Identity has always been the domain of Jörg Heuer Deutsche Telekom Laboratories – whose work I follow with interest and who was also a speaker at this event.

Speaking of IMS stack, Stefan Holtel. Vodafone Group R&D – D Munich raised the question of an Open IMS. He also questioned why IMS is not a part of the LAMP stack i.e. if we want to attract internet developers, then presumably we should be a part of the LAMP stack and open? The unfortunate situation today is: we see IMS developers almost exclusively confined to the vendor community .. Which is sad .. And which defeats the purpose.

Finally, there is the question of end to end IMS and SDP. End to End IMS is not an easy proposition since both Operators need to have IMS .. However, ironically many applications need end to end IMS(video calling). Hence, SDP may provide some elements of end to end working for IMS ..

More on all this soon ..

Any questions/comments welcome


  1. Peter Weik says:

    Hi Ajit,
    just one addition to IMS not being part of the LAMP stack. The IMS open source projects are there, it is more the problem that not too many developers outside the corporate domain want to have to deal with IMS neither do they seem to be to fond of it since it is always pictured as a walled garden approach.
    Together with my colleague Dragos Vingarzan I was also speaking at the FOKUS event that was mentioned in the post. We gave a tutorial and a workshop on the topic of the Open IMS Core project (www.openimscore.org) which brings the possibility to attach many SIP/IMS clients to a SIP or IMS SDP over IMS core components. It definitely helps IMS developers around the world already by today (just as projects like Sailfin, Mobicents, OpenXCAP or others do). So it is already possible to build an IMS network with clients, core elements, media gateways, enablers and SDPs on open source software, it is just not well exploited nor really relevant to service developers since none of the big telco operators dared to move forward by deploying IMS or open third party interfaces (yet ?).
    - Peter