By: Wedge Greene, Trevor Hayes

It is taking a long time for IMS (IP Multimedia Subsystem) to get uniformly delivered as part of all networks. After all, its original vision back in the late nineties was to provide one, interoperable carrier grade approach for providing all IP services. For its first leg of the journey, it was adopted by the mobile community and became a mechanism for 3GPP.

IMS is a bundle of specifications and standards developed by the telecommunications community to provide a consistent way of delivering and managing IP real-time communication services: initially voice and video calling, but always with digital convergence as the end goal. IMS enables interlinking any media server to any service device. From the outset IMS was based on a telco’s perspective of IP service delivery, incorporating traditional telco standards of security and management. IMS would be delivered via both protocol adoptions [mostly from the IETF] and interface standardization. It was to be a “standard” – in that everyone in the mobile community would use it to deliver component pieces and networks that interwork together.

A Typically Long Road

So IMS grew and evolved in the traditional standards-driven way: painstaking meetings to discuss specifications, voting, working out agreements among representatives of telcos and equipment vendors. Multiple vendors developed commercial products in a piecemeal way. Products were interlinked (often loosely) via the IMS-defined APIs. IMS implementation became a required, integral part of the purchase of mobile network technology. Further, convergence argues that as mobile demands IMS, so all networks should use IMS, and doing so become transport-agnostic. This makes deep sense to carriers.

But IMS had its critics who thought that the whole IMS edifice was too big, ungainly and complicated. Its anti-agile, long-winded development model made no sense to people familiar with the lean approaches being adopted by IETF. The rapidly growing Over-the-Top service providers knew that none of this was needed to deliver services over IP Networks.

They have a point. In 2016 IMS is still moving towards maturity. Its first big new service, VoLTE, began commercial inroads in 2015. That’s a long time in Internet terms: Google is about the same age as IMS. Facebook and Twitter are much younger. A myriad of OTT service providers deliver useful functionality over the Internet, and most of them are happily doing that without IMS.

Relevancy

So, has IMS taken so long that it has become irrelevant? Or is IMS alive and well, and offering new possibilities in IP convergence. Some old views of IMS are wrong. IMS, while not simplistic, is not as heavy and cumbersome a standard as many thought it might become. It isn’t a rigid prescription: IMS is a collection of APIs between functionally-defined, coarse-grained support services that can be used, or not used, in any specific implementation.

IMS plays an essential part in VoLTE, the mobile high definition voice specification that will replace circuit voice in the last mobile mile with VoIP. VoLTE shows the way, and not just for mobile, to eliminate the multiple IP-to-circuit-to-IP conversions needed today. As more carriers deploy and sort out interoperability problems, VoLTE calls will be IP end-to-end. VoLTE users will benefit from high-definition audio and video, seamless call handover between IP networks of any type, and a range of other IP services including presence, simultaneous voice/message text and file-sharing. Carriers will scrap lots of circuit-based technology and run more efficient all-IP networks.

But so far VoLTE ends up delivering services that look a lot like services that have been delivered by OTTers for years. After all, Skype arguably provides a lot of the functionality we will get from Voice-over-LTE, and it already works using a smartphone connected to WiFi.

Further, a key ingredient of IMS, and one that is fundamental to both Voice and Video over LTE, is Session Initiation Protocol (SIP). SIP can, and often is, deployed without the rest of IMS. SIP is used by some OTT VoIP service providers in component support aggregations of other IETF services. No IMS involved.