Recently I had a chat with Dean Bubbley about WebRTC & VoLTE on LinkedIn where you can find the whole thread. I put just the most interesting parts here, read to the end.:
Dean Bubley
Founder, Disruptive Analysis and Wireless Consultant
Sami, there are two separate problems to solve. They are assumed to be linked, but that may be a poor choice.
1) There needs to be a plain-vanilla “phone service” for LTE that works at least as well as GSM, and is ideally simpler. That could have been a precursor to VoLTE if it had been developed in time (5-6 years ago), rather than launching 4 years after LTE data networks. The problem is that the window is now almost-closed, as we are at or past “peak telephony” now. That makes the investment case much harder as few CFOs want to spend money for a legacy service with a declining market. Improvements in CSFB mean VoLTE is now optional for this use-case, except for CDMA operators. Arguably, a “velcro” solution keeping old GSM900 running for high-quality & reliability telephony is an option too, with dual-radio phones.
2) There needs to be a platform for “future voice & comms” services & applications. This needs to be open to developers to embed comms directly into applications; it needs to support new models of user interaction beyond traditional “calls”, and it needs to break the link between access subscription and service/application participation. In some cases they may be bundled, in others a Skype-type model will make more sense, or numerous other non-subscription approaches. In some cases an E164 number may be appropriate, in others it will be another identifier (or none). In some use-cases interoperability will be a benefit, in others it will be unnecessary or actually a disadvantage. Depending on the use-case, it may not need emergency calls, or fallback to 2G, because the application won’t recognisably be a “phone call”. It needs to be engineered around “intent & purpose”, not “minutes”.
The problem is that VoLTE falls between these two. It is desperately late, which means it’s losing the argument for Problem #1 except as a “distress purchase” by operators with no other choices. And it is nowhere near flexible-enough to be the basis for solving Problem #2 because of its 3GPP & GSMA underpinnings.
I had this discussion with GSMA years ago – they refused to start from the basic question of “What is voice comms going to be in the future?” and decided instead to recreate Telephony v1.1 and hope to add stuff to it later. Unfortunately that locked in both IMS as the platform, and the legacy “federation” model of the telecoms industry where each network implements its own local application stack.
Skype is a possible answer to Problem #1, yes, as is FaceTime or Viber. It “works” and can interop with the PSTN if needed. However none of these has the developer accessibility to be good answers for #2 either. (Skype used to, but has largely killed that as it’s being used by MS to add value to its own apps like Outlook & XBox first).
Some form of evolved WebRTC-type cloud platform seems most probable as a mid-term solution here. Although default will be OTT-style, it could even be coupled to network-based QoS functions if needed – after all, the PCRF guys keep saying their products enable “prioritisation” and guaranteed QoE for *any* 3rd party app as long as regulations permit.
If you want more detail on this, there’s a lot in my various Slideshares & blog posts, or my Future of Voice workshops & research reports.
——————-
Jan Sladky
IMS VoLTE RCS consultant
1.plain-vanilla “phone service” for LTE that works at least as well as GSM. Very nice idea.
I’m just curious about:
- who is going to invent it
- how can it become a standard ( no isolated islands please 
- won’t it be the same as VoLTE at the end ?
I have nothing against it, but my point is following – If plain-vanilla phone service should support following:
- to be reliable
- carrier grade quality (no mobile Skype please)
- emergency calls
- handovers PS-CS core and back, ideally with with video,
- supplementary services
- number portability
- interconnection with CS/PSTN networks
- legal intercept
then would not it be the same as VoLTE ? Because I think VoLTE follows the law of minimal energy – if you want all of the above features there is nothing simpler then VoLTE. We can talk about it for weeks but I’m almost 100% sure we would not figure out anything simpler that would be able to provide same quality and user experience as the GSM. Only VoLTE can fulfill all these requirements in reasonable time.
VoLTE is actually cool stuff, well standardized, offering much more then just the above list (ICS, roaming, etc.) It is undoubtedly the industry standard as the GSM Phase 1 was 20 years ago. And why the GSM beated the rest of the world like DAMPS, iDEN and CDMA networks ? because it was well standardized…how surprising
Yes, there is and there will be plenty of SIP IOT and other issues, but I believe it is matter of time to get that solved in reasonable time.
There is lot of interesting facts about VoLTE not known yet, e.g.
VoLTE handles calls much more efficiently the OTT in term of radio resources. The OTT vocie just simply needs more part of the radio channel than VoLTE. Normally it is 2-3 times more, under bad radio conditions it is 5x times more…User experince is of cource better with …. VoLTE codecs. Radio guys – keep it in your minds when deciding between VoLTE and Skype in your network
that’s my humble view. I might be wrong of course – I just believe in certain things with certain probability so there is always room to be wrong
I think VoLTE/IMS is flexible enough to solve the problem 2. I would be very surprised if something with such limited capabilities like WebRTC would have success in mobile world. But you now, Never say Never.
just for fun – my view on telco in 2020:
• network consisting of VoLTE/IMS + EPS + reduced CS,
• eSRVCC=YES
• ICS=YES
• legacy IN services migrated to IMS-AS
• VoLTE is SIMPLE but carrier grade, with all the telco
• WebRTC gateways connected to IMS but WebRTC not mass market mobile service.
I bet on VoLTE
——————-
Dean Bubley
Founder, Disruptive Analysis and Wireless Consultant
Jan
There are already more mobile WebRTC users than there are VoLTE users. That gap will widen explosively – although to be fair, most of those will not be directly-comparable “boring old telephony” calls, but other more interesting and rich use-cases, some data-only. I’m forecasting 1.5bn+ mobile WebRTC users by 2017, against perhaps 200m VoLTE users.
There are also more FaceTime Audio, Viber and Mobile Skype users on LTE than VoLTE users, and they & others will at keep pace if not exceed VoLTE in future as well. FaceTime Audio in particular can expect to be used as primary-telephony on future LTE iPhones, if Apple iMessage is anything to go by. Maybe Apple will use VoLTE as the fallback for FT Audio in the same way that SMS is the fallback for iMessage.
There are other several flaws with your assertion, not the least how any of that investment will be possible as revenues & profits from “vanilla voice” trend towards zero. Saving a little bit of spectrum from voice calls (nice in theory, hard to prove in practice) is quite literally going to be “lost in the noise” of other data traffic growth & declining call traffic anyway. Some more will be saved by virtualised or completely outsourced IMS as we move to SDN/NFV; open-source platforms like Project Clearwater might help.
I think in 2020 we’ll still see a lot of CSFB, especially as it is itself evolving towards lower setup latency. We *might* see a return of VoLGA-type models tunneling CS calls over LTE, or dual-radio GSM+LTE phones, but both of those (very sensible) models appear to have been killed by IMS-centric telco fundamentalists. A pity, they both might have saved lots of telcos.
The other thing we’ll see by 2020 is a large proportion of voice (& video) communications not being done as standalone “calls” or billed services, but embedded directly into apps and websites as “features”, often mediated via WebRTC and 3rd-party API/cloud players like Tropo, Twilio, Temasys, Tokbox (Telefonica) & a dozen others. Some of these *might* be QoS-enhanced via telcos’ own network APIs and policy-infrastructure, but that will be the exception rather than the norm. Almost none of the “raw ingredients” for embedded communications will be IMS/VoLTE services exposed via APIs & gateways.
——————-
Jan Sladky
IMS VoLTE RCS consultant
Hi Dean,
1. WebRTC surely has its place in communications, there is no question about it. But to compare just number of current users has very small informative value. It is like if you compare number of bikes with number of Volvo XC70. I like my bike and I use it a lot during summer. But I rely much more on my Volvo XC70, because:
a) if I switch my well designed Volvo engine on (i.e. VoLTE phone) the car is running and all the necessary sw is ready, e.g. notification about lack of cooling liquid. On the other hand if you switch the OS on, you still need to launch the web browser to be available. Very small detail, but having big impact. (I’m not talking what happens if you close the browser and someone wants to call you
b) Volvo XC70 can protect me against rain, snow, wind, etc.: VoLTE protects the sound quality via the dedicated bearer – the sound quality is superb. WebRTC doesn’t even have echo cancellation, sound quality is poor. Have you ever tried ? I did.
c) Volvo XC70 can also protect me in case of accident: VoLTE has emergency calls as standard, WebRTC will most likely never have it. (Or does Skype have it after 10 years ? Of course not).
d) Volvo XC70 as AWD/4×4/all terrain car. I do not care whether I’m on highway between Pilsen and Prague or somewhere off-road in Sierra de Guadarrama, Spain. I simply know that I always reach the destination. VoLTE is the same – you don’t care in which domain you are – CS core, PS core or whether you are moving from PS to CS or vice versa. You always can make a call and you always reach the B party. On the other hand, with WebRTC you rely on things you have no control on – network throughput, latency, etc. and you just pray you will hear each other.
d) With Volvo XC70 you have your 240 horsepower engine. With VoLTE you have your native phone application, embedded, native VoLTE UI. With WebRTC you just have interpreted Javascript or in the best case the HTML 5 which both must be interpreted by the browser engine. My guess is that it is not going to be rare case to see frozen browsers when calling, downloading or browsing.
You say “WebRTC will bring other more interesting and rich use-cases”. Ok, but what it is going to be ? I guess it is not going anything special … Especially compared to the VoLTE and RCS on the same device.
On telco market, there is room for everything good and reasonable, even for WeRTC, for sure. WebRTC on Facebook to call your buddies will certainly be cool for teenagers, similarly on web of you insurance company the WebRTC will let you dial 800 quickly.
But with the simplicity behind the WebRTC it is hard to believe, it will become a important mobile phone application – but who knows, time will tell.
My experience is that WebRTC is currently more Tin Can Telephony (http://en.wikipedia.org/wiki/Tin_can_telephone) than something serious. But I do not underestimated it – can evolve quickly, the Web is ubiquitous.
BTW is that true that Microsoft submitted their own WebRTC proposal to the W3C that is different to the rest ? And how it looks like with Apple and his Safari – still no support ? Is also true that there is several WebRTC frameworks that are not compatible ? Maybe these poor guys from Microsoft, Google, Apple and Open Source community would need something like GSMA No, I’m just kidding, they certainly know what they are doing – the best will rule the WebRTC world.
Summary: Sometimes it is ok to drive a bike, but there are also situation when you would rather switch to Volvo XC70 like comfort and capabilities – and that’s what VoLTE is actually about. VoLVO=VoLTE.
Don’t forget – I just believe in certain things with certain probability so there is always room to be wrong.
regards, Jan