Will IMS Ever be Used to Create Services and Process Media?

August 26, 2014

The quick answer? Not really.


I have decided to write a quick post here to a question I bumped into on Quora: What would happen to media gateways and gateway controllers after communication gets unified

My answer on Quora?

The whole notion of separation between the media gateway and its controller is a bit quaint to me.

In most cases, I see the separation as an architectural one that is internal to a product than it is something that can be integrated across different products.
A controller of a gateway has other qualities then the media gateway itself – especially in the domain of how the two entities scale, but should they be separate in the sense that one is built by company A and the other by company B?

I think both will come from the same vendor.

It got me thinking a bit further – the separation of gateways and gateway controllers was something that has been designed in the past already. There are protocols used to communicate between the two separated entities. This has grown in complexity when it got introduced into IMS.

From there, it was a simple for me to shift over to the discussion of IMS. People in love with IMS tell me at conferences that there are already IMS deployments out there, indicating that carriers should continue working on IMS towards building services on top of it.

Here’s my view: IMS is here to stay. But only in the backend systems of the operators to play policy and billing to some extent. And in places where it replaces old non-IP services (voice and SMS), and even that is questionable for some.

The notion goes today that VoLTE is upon us. If we just give it a slight nudge, a bit of extra time. We will all see IMS at its glory. My problem with it, that the same tune has been playing for about a decade. VoLTE will happen, but not because it enables the creation of new services, but rather because it enables carriers to power down old systems in favor of all-IP ones.

Processing media in the server side? Creating rich new services in no time? All these just can’t happen in an IMS world.

IMS got over engineered. It became impossible to build with, especially when you think of the available alternatives.

As we migrate to IP networks, anyone can build a service – it now isn’t locked down and controlled by the carriers. OTT players can now do everything a carrier can when it comes to introducing a brand new service. So why bother with an architecture from 10 years ago called IMS when we learned so many new and interesting things about the web and scaling in the meantime?

If you ask me, services won’t be created in IMS. It will be created over proprietary infrastructures that are web based in their nature and use open APIs to cultivate collaboration and creativity.

You may also like

Fixing packet loss in WebRTC

Fixing packet loss in WebRTC

Your email address will not be published. Required fields are marked

  1. IMS does not bring services. It is a session control layer allowing services via a Service Layer built with AS such as SDP or API Gateway. It offers a service to be notified of any event from the user or the network. It allows messaging, session…
    It also facilitate the interworking with other operators, legacy network, various UE…
    It finally bring features such as charging, legal interception, data modelling…

    So IMS for sure alone does not bring services except the basic services such a identity presentation, diversion, conferencing…

    1. Crutel,

      The only point is that most (all?) services won’t be created on top of IMS, but rather on top of other types of infrastructure. These will connect and integrate with IMS only when truly necessary.

      So the session control layer you speak about? Useless to anyone but a telco launching his own basic services (even telcos will end up launching a lot of their services outside of the IMS session control layer).

{"email":"Email address invalid","url":"Website address invalid","required":"Required field missing"}