The IP multimedia subsystem architecture concept has captivated the telecom industry with the promise of becoming a next-generation network core that will lower traditional barriers between access mediums and support efficient, seamless delivery of convergent services. But, as with any much-hyped technology, there has been a lack of discussion of the specifics of IMS, particularly its impact at the network element level. Telephony's Dan O'Shea recently talked to Grant Henderson, co-founder and executive vice president of marketing and strategy for Convedia, a company that provides media processing functions within IMS architecture, about the functional implications of IMS and the future of this potentially industry-altering concept.

On the hype surrounding IMS: What is IMS, apart from being a collection of standards? Well, it's been embraced by the industry to mean pretty much anything and is being used so much, it's in danger of losing its meaning. Some people say IMS is a fixed/mobile convergence (FMC) concept, and some people say it is more about VoIP spreading through the network. It's a bit of both because the call session control function in an IMS architecture is nothing more than what a softswitch is in a VoIP deployment.

On IMS network elements: Softswitches weren't designed to do the mobility hand-off, but we're in the earliest phases of IMS, and you are going to see different elements evolve elegantly to support IMS. All-singing, all-dancing IMS is still a couple years away for most carriers. The first step in migrating to IMS is to get that softswitch core built out and begin bringing IP traffic onto that core. What IMS does is put a stick of dynamite in the old Class 5 switch. It changes the way services are developed in the network to be more fluid and less monolithic. You are going from a silo approach to services, to a more layered approach, and it's all based on IP MPLS technology. One of the things that's different about an IMS architecture is the host subscriber system. Traditional elements like media servers and application servers correspond to similar elements in IMS.

On Convedia's role in IMS architectures: Media servers are core network elements in VoIP, IMS and FMC deployment. An IMS deployment without media servers would be somewhat neutered. In an IMS deployment, the media server is what's known as a media resource function processor. It contains all the media and application server logic, and it provides the network-based processing of media for complex applications like push-to-talk. BT currently is in the process of moving to its 21st Century network — changing out its old [public network] architecture. In that network, our media server will be controlled by a softswitch.

On what carriers expect from their IMS vendors: When it comes to IMS, the Tier 1 operators are looking for large vendors. That could lead to some consolidation of some of the smaller players, but basically, they want their vendors to be systems integrators. There will still be room for best-of-breed solutions because these elements will be based on open standards like SIP.

On FMC: IMS is not just about FMC. You see carriers going into FMC for different reasons. I think when you saw the economic bubble burst in the fixed-line market a few years ago, the wireless guys were not as harshly affected. Everyone was watching the BT Fusion FMC service, which is based on the Unlicensed Mobile Access (UMA) standard. Whether or not other carriers follow BT in doing an FMC service will depend a lot on the capabilities of mobile/Wi-Fi handsets and how those shape up. Carriers also still face challenges in learning how to market enhanced services. If FMC fails to take hold as a service for some reason, it could slow our network evolution, but the constraints in that case would be limited to the access network. Even if that happens, service providers will still adopt IP at the same rate because they are using it to drive costs out of their networks, regardless of FMC or other applications.

On IMS services and handsets: If you look at something like UMA, there has been some questioning of how many handsets would be ready and how those handsets would perform. Not everyone is going to evolve to UMA, but it now looks like a sizable number of handset vendors will support it. With IMS, you may see something more similar to how the VoIP market has developed. A lot of VoIP service is still enabled to old access phones. You will see a more gradual evolution, an evolution in steps. We are talking about voice, but video, messaging, personalization, conferencing and other applications will be affected.