The timing of the VoLTE introduction is tied to a) the testing programme b) device support at chip level for SRVCC c) service availability on 800MHz spectrum that will give EE an even wider reach for LTE (there’s not much point having VoLTE if you’re going to be handing in and out of coverage all the time).
Speaking to journalists and bloggers taken on a trip round EE’s test centre north of London, Tom Bennett, Director of Network Services, EE, said, “This is the most complex thing I’ve ever worked on in 20 years.”
“With most services the difficult thing is getting something working the same way over and again. With VoLTE, getting it working once has been a pain,” Bennett added. EE has 50 of its own engineers testing the service, and will later extend that to 450 customers connected to its test systems.
Bennett claimed that the introduction of LTE itself required seven new “systems” – referring to the EPC-SAE architecture mandated by LTE – whereas VoLTE has required 17 new “systems” and 74 new interfaces, he claimed.
This is a reference to the introduction of the IMS – and the supporting signalling and control resources – required to support VoLTE. From what I could see, in the test centre EE is running a Mavenir MRF – I saw both COTS-based on HP servers and ATCA Blade – and a Huawei TAS (Telephony Application Server). Not sure about the rest of its IMS to date.
Journalists were shown both a VoLTE call and a VoWiFi call between two devices over live networks, supported by EE’s test IMS.
In the subsequent Q&A it was instructive to see the reaction of journalists who have not been exposed to the world of IMS and VoLTE before come to grips with it.
So, asked one, if it’s so complex why are you doing it – what are the advantages to the consumer? Lower set up times and better quality, came the reply – although quality is similar to 3G HD Voice (it’s the same codec) at the moment, so the quality uplift is more based around future improvements to the codec.
What about revenue opportunities? No, there won’t be any revenue increase from VoLTE, although there may be some future IMS-based services that may be revenue additive.
Why is it so complex? Didn’t you [the industry] know this would happen when LTE was designed?
“Well, when LTE was specified, everyone thought IMS would long be deployed by now,” came the answer.
So you could see the wheels going round in the heads: it’s very complex – therefore expensive, many of your current devices can’t support it, it’s not going to make you any more money and you’re not going to market it as a service in itself.
(And these questions were without EE telling them that devices are going to need SRVCC (single radio voice call continuity) support at the chip level (so not just a firmware upgrade))
So why do it?
The advantages lie mostly with the operator, EE admitted. In fact, the proposed benefits to operators of VoLTE are well-rehearsed: with VoLTE ubiquity they can eventually switch off 2G/3G services gaining much higher spectral efficiencies from existing spectrum, they can benefit (perhaps) from introducing new revenue-generating services over the IMS, and there are potential operational savings from switching to all IP instead of a CS core. For more of those operator benefits, see this eBook produced by TMN in association with Radisys (incidentally, Radisys is the platform provider for Mavenir, whose kit EE is using).
Put simply, operators sort of have to do it – they can’t have calls falling back to 3G all the time, it’s just too annoying for consumers and signalling intensive. So they need to support VoLTE – and to do that they need the IMS. The justification for this investment is that as voice is a key service (900 million calls a week on EE’s network from its 27 million customers) it’s worth investing in, even if it won’t actually increase your top line. It’s a slightly confusing world where you need an IMS because you need VoLTE, and you need VoLTE to justify your investment in the IMS.
EE looks likely to be another that will announce support for WiFi calling on IoS8 devices later this year.
EE’s decision to use its infrastructure to support VoWiFi will cheer up one company – Kineto Wireless. Kineto is already supporting Sprint’s VoWiFi service and is keen to shout about this tangible consumer advantage of the IMS – supporting WiFi calling from a device without the requirement for an app or separate number.
And with Apple’s IoS8 supporting VoWiFi (before VoLTE) operators have another compelling consumer reason to support native WiFi calling, with Rogers Wireless and T-Mobile already confirming they would support the service. EE looks likely to be another that will announce support for WiFi calling on IoS8 later this year.
A minor irony, then, that one of the early benefits of introducing the supporting infrastructure for VoLTE, a technology that will work best with widespread LTE coverage, is that operators can make use of a technology where there is no coverage.