ABI Research claims there is increasing spending on all the core network elements required for a functioning IMS network as telcos look to support their LTE deployments with the packet voice standard - VoLTE (Voice over LTE).

According to the firm the annual IMS market for those core products will reach US$4 billion by 2017, after which the revenues will flatten out as that particular wave of investment ends.

But will it... really? Might it not be that IMS has had its window shut and what we’re now seeing is not a slow edging up but a firm edging out.

Even if carriers do decide to continue on a with a ‘standard’ IMS approach to future voice support, the chances are that they will adopt various flavours of virtualisation to radically reduce costs - an option also acknowleged by ABI Research in its report.

In its raw state LTE offers mobile broadband services only, unlike 2/3G networks and services which offer both switched voice channels and data channels. As a result most telcos are adding their LTE mobile broadband atop their existing 2G/3G networks (unless they are greenfield challengers) with the idea of ‘falling back’ to switched voice on the earlier networks when a voice call is to be made.

Eventually, though, it’s anticipated that operators will likely want to retire their older network technology and ‘refarm’ their 2G spectrum for use by LTE, at which point they will need a supporting technology for voice over packet. This is where VoLTE and IMS comes in. IMS (IP Multimedia Subsystem) is the standardised call-handling framework for the packet network and VoLTE is the voice over packet over LTE format for the mobile device.

Now at this point things could get very disruptive because the goal-posts have all been moved since work started on IMS. It was originally envisaged that as well as voice, IMS would be the basis for all sorts of enhanced communications services (hence the ‘multimedia’) such as those developed under the RCS (Rich Communication Services) banner.

This is now looking far less likely. In fact (be bolder Ian) this is now acknowledged as a non-starter. Things have clearly moved on and new communications services look certain to be driven by standard IP whether they be launched by so-called OTT players or by the carriers themselves. No IMS required.

The question then becomes: if voice is pretty-much the only candidate for IMS, is IMS still worth supporting? Might it not be better and more forward-looking to support other - less complex - technology flavours for VoIP instead?

If the answer is still IMS, but IMS virtualised and running on bare metal servers and made available to carriers at a fraction the cost of the original, non virtualised black box network components, how much of that US$4 billion will telcos really still need to spend by 2017?