Operators Raise Voice Services on LTE

COMMENTS

Newest First | Oldest First | Threaded View

<< < Page 2 / 2

Dean Bubley 12/5/2012 | 4:47:43 PM

re: Operators Raise Voice Services on LTE

I'm somewhat amazed that nobody has yet commented on the obvious solution - and one which is (quietly) gaining support as a backup.

That is the notion of using a version of SVLTE but for simultaneous dual-radio support of both LTE and GSM/UMTS. So just as in CDMA smartphones, where we get concurrent data-on-LTE + voice-on-CDMA, we'd move to the equivalent for 3GPP networks.

Dual-radio LTE+GSM/UMTS makes a huge amount of sense, and discussions I've had suggest that various device & silicon vendors have it as an option.

The argument about battery life won't wash - you could partition off 200MAh which is probably enough to power an efficient GSM phone for a week. More tricky is the registration (you might need two SIMs or 2 IMSIs) but that could end up being a feature rather than a problem.

The problems of CS Fallback are not restricted to the call setup time (which in some circumstances can add >2 secs I've heard - tell that to your emergency-calling legal team). The killer is the fact that the data connection drops when you get or make a phone call, possibly (if you're lucky and in coverage) reconnecting on 2G/3G.

That is such an old-school view of the world (ie telephony is *always* the most important thing you'd do on a "phone") that it's almost prehistoric. So the CFO signing off the end-of-quarter spreadsheets on a mobile cloud-based version of SAP loses the connection when he gets a call from a doubleglazing salesman? Or the kid downloading a (paid!) video or app gets it interrupted when a parent calls, and then calls customer service to make sure he wasn't charged twice?

In an era of multitasking phones & people, it seems incredible that we're deploying a brand-new technology intended for single-tasking.

As a side-effect, it will make it *so* much easier for people to justify using Skype or Viber on LTE for their most important contacts, just dropping down to circuit voice as a lowest-common denominator when they absolutely need to.

VoLTE will probably be the long-term solution for some (but definitely not all) LTE operators, although I have my doubts about the Q1 2013 timeframe. I reckon it'll be 2015-2016 before it's fully robust enough for massmarket deployment.

Plenty of other LTE operators will either not do a dedicated telephony service at all, or will deploy another version of SIP-based NGN VoIP, or just partner with Microsoft/Skype, Google or whoever. We may even see HTML5 browsers with WebRTC maturing quickly enough to support speech via the web by then.

All these are topics discussed in my Future of Voice workshops, run with Martin Geddes. www.futureofcomms.com

Dean Bubley

Disruptive Analysis

Reply | Post Message | MESSAGES LIST | START A BOARD

Gabriel Brown 12/5/2012 | 4:47:42 PM

re: Operators Raise Voice Services on LTE

@Dean Bubley That horse has bolted.... we can debate the reasons, but the horse has dissappeared over the hill.

Also, per your example, CSFB should do an okay job. Most data apps are resistent to a short delay on handover (or even cell reselection for that matter), so in practice it should be good enough for the near-term. The CFO cloud example is a bit extreme, but is perhaps a good reason why operators should go for the full VoLTE machinery. 

Reply | Post Message | MESSAGES LIST | START A BOARD

Gabriel Brown 12/5/2012 | 4:47:42 PM

re: Operators Raise Voice Services on LTE

@Nietzsche Agree. That pretty much captures why VoLGA didn't fly.

Reply | Post Message | MESSAGES LIST | START A BOARD

Dan Warren 12/5/2012 | 4:47:41 PM

re: Operators Raise Voice Services on LTE

Much has been said about why UMA/VoLGA lost to CSFB as an interim solution and VoLTE/IMS as the long term goal.  Service parity is one explanation (but then CSFB is not exactly allowing anyone to break out of the mold).

The main reason was timing.  There was an assumption in 3GPP that IMS and MMTel would be the basis of the long term solution, and that was why CS bearers were dropped from LTE specs.  That assumption was only crystalised when the 'One Voice' initiative appeared late in 2009, and that work was then taken up in GSMA as the basis of VoLTE.

In the meantime, the industry saw a potential gap that needed to be filled by an interim solution and CSFB was defined to fill that gap.  After that, VoLGA was proposed as a Study Item in 3GPP, but when that study reached conclusion and was about to be turned into something that would result in specs, 3GPP began to see the issues that could come about with having three different voice solutions (plus existing CS on 2G/3G) in the market place.  Roaming primarily would become a pretty apocalyptic mess.  LTE was being cast as the technology to unify the industry, and yet we were in danger of ending up with voice call roaming and interconnect becoming more difficult than ever before.  As a result, VoLGA never became a 3GPP spec in order to keep the overall voice service as simple as possible.  Ultimately, VoLGA was a diversification too far.  The companies backing VoLGA formed the VoLGA Forum, which produced their own specifications, but what was lacking was operator support - DT stuck with it for a while, but pretty much backed out in the first half of this year.

As I have mentioned previously, in a technically purist world, I would push very hard for CSFB to not happen either, but that is not what the commercial requirements dictate.  Voice with LTE is needed now (even if the voice service isn't really realised in LTE directly) - for some markets, technologies live or die on the availability of handheld devices, and the consumer expectation is that those devices have a voice service that is at least as good as that which they have today from CS.  An IMS based future is where the mold could be broken, but it relieson the operators to embrace it to the fullest and the vendors to implement it - we are in the usual chicken-and-egg situation of one side saying 'we'll commit to develop if you guarantee volumes' and the other saying 'when you can show us the product, we'll commit to deploy'. At some point (around the end of 2012) we tip over the edge of that process and the cycle becomes positive and ecosystem-generating.

 

Reply | Post Message | MESSAGES LIST | START A BOARD

<< < Page 2 / 2