I listened in on a call yesterday that discussed the ongoing issue of voice and SMS on LTE networks. I've written before about 3GPP and certain vendors trying to use LTE as a lever to strong-arm operators into adopting IMS.
A year ago, it seemed like IMS had finally been put out to pasture by the mobile industry. Some innovation was occurring with partnerships like 3/Skype and E-Mobile/Jajah and Mobilelom/fring. But I underestimated the perpetual and desperate attempts by vendors, the 3GPP and the GSMA (through the ill-thought-out RCS initiative) to position it as the "default" choice for services on LTE networks. Although some vendor representatives grudgingly admit that LTE doesn't *mandate* IMS, it's still being heavily skewed towards it.
I've long held that IMS voice is precisely the wrong choice for mobile networks, for assorted reasons relating to handsets, web integration, emphasis on "multimedia", dependency on SIM cards and the overall cost/complexity involved. I'd now go further and say that the whole notion that "sessions" are a good way to describe human-to-human communication is fundamentally wrong - look at social networks as an example of this.
But despite the efforts of certain vendors to embrace the inevitability of integration with Internet applications, others still seen unable to grasp the realities. They can't bring themselves to mention words like "Skype" or "Facebook", "iPhone" or (whisper it) "Google Voice and Google Wave". They can't even bring themselves to use words like "handset" or "smartphone", still referring to "terminals" as if they were dumb green-screen boxes from the 1980s.
I'm simply staggered that some people still brazenly claim that IMS is "inevitable" in mobile, using phrases like "platform of choice". It just defies reality, especially from an end-user perspective.
Initially, I was going to make an ostrich analogy. But then another bird seemed more appropriate....
(With apologies to John Cleese and Michael Palin)
Scene: The Mobile Infrastructure shop. An operator enters a vendor's shop.
Operator: 'Ello, I wish to register a complaint.
(The vendor does not respond.)
O: 'Ello, Miss?
Vendor: What do you mean "miss"?
O: I'm sorry, I have a cold. I wish to make a complaint!
V: We're closin' for lunch.
O: Never mind that, my lad. I wish to complain about this IMS platform I purchased not half an hour ago from this very boutique.
V: Oh yes, the, uh, the 3GPP standard...What's,uh...What's wrong with it?
O: I'll tell you what's wrong with it, my lad. 'E's dead, that's what's wrong with it!
V: No, no, 'e's uh,...he's resting.
O: Look, matey, I know a dead application platform when I see one, and I'm looking at one right now.
V: No no he's not dead, he's, he's restin'! Remarkable platform, the 3GPP standard, idn'it, ay? Beautiful combinational multimedia services!
O: The multimedia don't enter into it. It's stone dead.
V: Nononono, no, no! 'E's resting!
O: All right then, if he's restin', I'll wake him up!
(shouting at the cage)
'Ello, Mister IMS! I've got a lovely fresh LTE network for you if you show...(owner hits the cage)
V: There, he moved!
O: No, he didn't, that was you hitting the cage!
V: I never!!
O: Yes, you did!
V: I never, never did anything...
O: (yelling and hitting the cage repeatedly) 'ELLO IMS!!!!!
IMS!!!! Come on, wake up! IMS!!!
(Takes application platform of the cage and thumps its VoIP and messaging servers on the counter. Throws it up in the air and watches it plummet to the floor.)
O: Now that's what I call a dead platform
V: No, no.....No, 'e's stunned!
O: STUNNED?!?
V: Yeah! You stunned him, just as he was wakin' up! IMS platforms stun easily, major.
O: Um...now look...now look, mate, I've 'ad just about enough of this. That platform is definitely deceased, and when I purchased it not 'alf an hour ago, you assured me that its total lack of movement was due to it bein' tired and shagged out following a prolonged wait for a suitable LTE network.
V: Well, he's...he's, ah...probably pining for the Rich Communications Suite
O: PININ' for the RCS?!?!?!? What kind of talk is that?, look, why did he completely ignore the Internet the moment I got 'im home?
V: The IMS prefers ignorin' the Internet! Remarkable platform, id'nit, squire? Lovely multimedia combinational services!
O: Look, I took the liberty of examining that platform I got it home, and I discovered the only reason that it had been sitting on its LTE network in the first place was that it had been NAILED there by the standards committee.
(pause)
V: Well, o'course it was nailed there! If I hadn't nailed that platform down, it would have nuzzled up to those OTTs, opened its APIs, and VOOM! Feeweeweewee!
O: "VOOM"?!? Mate, this platform wouldn't "voom" if you put four million sessions through it! 'E's bleedin' demised!
V: No no! 'E's pining!
O: 'E's not pinin'! 'E's passed on! This platform is no more! He has ceased to be! 'E's expired and gone to meet 'is maker! This is a late platform!
'E's a stiff! Bereft of life, 'e rests in peace! If you hadn't nailed 'im to the LTE network 'e'd be pushing up the daisies!
'Is metabolic processes are now 'istory! 'E's off the handset!
'E's kicked the bucket, 'e's shuffled off 'is mortal coil, run down the curtain and joined the bleedin' choir invisibile!!
THIS IS AN EX-PLATFORM!!
V: Well, I'd better replace it then
O: If you want to get anything done in this industry, you've got to complain until you're blue in the mouth
V: Er, sorry guv, but we're right out of platforms
O: I see, I see, I get the picture
V: I've got a DPI and differential QoS box
O: Does it work?
V: Not really, no
O: It's scarcely a replacement then, is it?
V: I tell you what.. if you go to my brother's infrastructure shop in the cloud, he'll replace your platform for you.
O: Alright
Dean Bubley's Disruptive Wireless: Thought-leading wireless industry analysis
36 comments:
:-)
So true.
Pffft, IMS has plenty of flaws but what alternative do you propose for the IP control plane of the future? There are already 100s of IMS implementations worldwide and I believe the momentum is pretty much unstoppable now.
There may be fixed VoIP implementations, or a very few isolated examples in mobile (or simply rebranded R4 MSC servers), but there's certainly no momentum behind the use of IMS for mainstream mobile applications.
There are certain control aspects for which specific IMS elements might make sense (eg policy), but the idea that future cellular voice / messaging etc will be delivered by MMTel and some mythical "combinational services" is a fantasy.
There is no way that modern communications platforms like Facebook could have been created in an IMS world: the whole notion of "sessions" for person-to-person communication now seems quaint and very 20th century.
I see a very long legacy of basic CS voice and SMS, blended to the web for most of the value-add and cool context-based modes of future communications.
That doesn't mean that mobile operators cannot add value, but they need to be much closer to Skype than to the IMS vision. Actually, come to think about it, maybe Vodafone or T-Mobile should acquire Skype....
Dean
XMPP is possibly usable for IP control plane. Google and Apple make extensive use of XMPP in their signalling.
iPhone push messages are XMPP messages send by XMPP's PubSub mechanism.
Google uses XMPP in the signalling plane of Wave and in GTalk.
This may be your opinion but it is not what I am hearing and seeing in telcos (mobile and fixed) around the world all of whom (mobile) are launching IMS based services with a view to IMS call control in the future (or for fixed operators - now). I suppose this could all come to a crashing halt as you suggest but I honestly think your contrarian opinion is just opinion and not borne out by the facts of the situation.
Also, the notion that facebook could not have been created in IMS is a red herring. Do you think that operators would implement email on IMS for example? Horses for courses, but there's nothing to stop a facebook like application using IMS services to provide communication services to its users which is more along the lines of how it is envisioned this kind of thing would work.
And what are these supposed IMS services that these operators are launching, and on what devices?
Push-to-talk is niche
Video-sharing is pointless
RCS is not fit-for-purpose
MMtel is unsuitable for VoIP
Presence & IM is a battery-killer
Hosted PBXs aren't massmarket
All have plenty of alternative more-usable substitutes available now. In particular, there's a boatload of web-based telephony APIs and platforms (using both CS and VoIP) that can be used today for mashups.
IMS has not been developed with an eye on user behavioural psychology, the realities of device evolution, or the role of 3rd party developers. It has baked-in assumptions about how people consume service and applications that are demonstrably false.
When I see Apple release a version of the iPhone software with a full IMS client, I'll believe it.
I've also heard recently that another major handset vendor has significantly reduced its IMS software engineering team - hardly a signal that it's going to be a mandatory capability in future.
It seems to me that your objection to IMS is based on the fact that nobody has cracked the mobile client (yet). I think you are missing the bigger picture: operators acknowledge that they need an IP based control plane for the future that is standards based (i.e. interoperable and supportable) and allows for convergence across different access types. As access, transport and control all move toward IP, what other option is there but IMS? We will still have an awful lot of GSM CS voice in 10-15 years time but I expect the control plane and service layer will be IMS.
IMS has plenty of flaws but large operators have already (rightly or wrongly) settled on it as the control plane of the future.
DB: "IMS has not been developed with an eye on user behavioural psychology, the realities of device evolution, or the role of 3rd party developers. It has baked-in assumptions about how people consume service and applications that are demonstrably false."
Are you suggesting these same problems are not inherent in CS based mobile networks?
> Are you suggesting these same problems are not inherent in CS based mobile networks?
Different situation - firstly, there's 100 years of learned consumer behaviour around telephony in general, and 10-20 years around CS mobile.
Secondly, if you were designing CS mobile today, in a "greenfield" scenario, there's plenty you'd do differently.
Thirdly, there's 10+ years of learned behaviour and expectations around IP-based communications generated by the Internet and other platforms.
Fourthly, there's 10+ years of experience of the breadth and pace of 3rd-party innovation around IP communications services... which is now also leaking over to CS via the various Web/Voice 2.0 platforms.
Anonymous >>It seems to me that your objection to IMS is based on the fact that nobody has cracked the mobile client (yet).
The problem is that they're getting further away, not closer. I first identified this as an issue in a report published in June 2006, and I'd been tracking it for more than a year prior to that.
Since then, we've had two cycles of Moore's Law, the increasing shift to smartphones, multi-access, full web browsers, Ajax, rich handset APIs, iPhone-grade UIs, widgets, appstores and plenty more.
It is now 10x harder to get IMS working on the device, because the client has to work with/around all this other stuff. In addition, the nature of the web & widgets in particular have made the notion of fixed "enablers" look quaint & 20th-century.
Anonymous >>I think you are missing the bigger picture: operators acknowledge that they need an IP based control plane for the future that is standards based (i.e. interoperable and supportable) and allows for convergence across different access types.
No, the biggest fallacy of all in IMS is this notion that apps should be "bearer agnostic". Increasingly, it's bearer-awareness that's becoming important. Not only that, but "control" is migrating to the edge, not concentrated in the network. There will need to be a collaborative process between the network, the application (which may reside partly on the web, through mashups), the device OS and sometimes the user.
IMS assumes that the operator *can* retain end-to-end control, when the evidence suggests this is both unrealistic and ultimately undesirable.
That doesn't mean that the "pipe" is dumb. There is plenty of value to be added. But it will be *dumber* than both the endpoints (handsets + GHz processors) and "The Cloud".
Anonymous>> As access, transport and control all move toward IP, what other option is there but IMS?
Last time I checked, both the Internet and corporate networks seem to run perfectly well without it.
>> We will still have an awful lot of GSM CS voice in 10-15 years time but I expect the control plane and service layer will be IMS.
Yes, it's possible that some elements of IMS may be used internally within operator networks for specific functions.
But now I think about it, I'm wondering if the entire concept of a unified "control plane" is itself obsolete. Control will inevitably cluster along with intelligence and computing capability, I think.
There will be some of this compute power in the operator network, for sure. But it will generally not be capable of fine-grained control.
We already see this in areas like DPI and QoS - all the rhetoric around tracing and billing individual "applications" has been shown to be unrealistic, because of much greater complexity higher up the stack.
>Push-to-talk is niche
>Video-sharing is pointless
>RCS is not fit-for-purpose
>MMtel is unsuitable for >VoIP
>Presence & IM is a battery-killer
>Hosted PBXs aren't massmarket
I agree that IMS has a very difficult business case. However, I think RCS is quite a good idea.
After all, phonebooks and sms menus in our phones have not changed in a decade. RCS is a way to put in a phone something similar to Skype-like interface.
Somethings seems to happen on RCS in France:
"Bouygues Telecom, Orange and SFR announce a user trial for new multimedia communication services for the second half of 2009"
http://www.francetelecom.com/en_EN/finance/news/cp090211en.jsp
Can you please elaborate more on why RCS is not fit-for-purpose?
Davide
My views on RCS are well-known. Numerous limitations, notably no out-of-the box Internet integration, no 3rd-party developer support, unclear model with prepay, exclusion of Apple and RIM from the initial (& current!) participants, presence kills the battery, undesirable link between phonebook & access contract [do we need contact portability laws?] and various others.
The French operators appear to be the main proponents of it at present, plus certain people at 3 or 4 other Tier-1 MNOs.
Various others are ambivalent or outright hostile.
My first comment was last year:
http://disruptivewireless.blogspot.com/2008/04/ims-rich-communications-suite-necessary.html
I'm also an associate of Telco 2.0 & concur with the views in this article (and see my own comment at the end)
http://www.telco2.net/blog/2009/02/rcs_really_considered_harmful.html
Some good comments here too:
http://blog.dial2do.com/2009/02/27/telco-20-guys-put-the-boot-into-ims-rcs/
One amusing thing about RCS was that for some time, a link to my blog post above was included on the Wikipedia page. Not by me, I don't know who put it there. But it's conspicuous that someone else edited it out a while later, after it had become one of my most-read blogposts....
Dean
Your points are interesting Dean, I just don't find them terribly convincing. Your argument that IMS is effectively a dead duck flies in the face of the many many millions that has been invested in it by every operator in the world of any size.
IMS is by no means perfect (and you make valid points as to its flaws) but it has been selected by the industry and as such has a firm future.
IMS being bearer agnostic does not imply that it has no bearer awareness, quite the opposite, IMS apps are capable of adjusting depending on the access type.
DB: "Last time I checked, both the Internet and corporate networks seem to run perfectly well without it. "
I know that you are being facetious but I'm not sure what you are implying here?
If I (as a big fixed & mobile operator - 20 million users in my case) want to migrate my network to IP everywhere: access, transport and control, what do you suggest I should do with the control layer? Whack something together with XMPP as somebody suggests above? Run my CS network until it falls apart? Just forget about services and leave them to 3rd parties altogether?
Another problem with RCS is that SIP is used for signalling. Presence using SIP has serious scalability problems when it comes to cross domain traffic. The problem with SIP is that messages are huge, must larger than presence messages in other protocols.
SIP is definitely useful in some cases but presence is not one of them.
> Your argument that IMS is effectively a dead duck flies in the face of the many many millions that has been invested in it by every operator in the world of any size.
These millions are peanuts compared to the investments that are necessary in the radio network.
Cheers,
Q
Anon: "These millions are peanuts compared to the investments that are necessary in the radio network."
Yes, perhaps (I think this will vary hugely depending on the age of the operator's network) but I would argue that this is investment that would have to happen for the most part irrespective of whether IMS is implemented or not.
>>IMS is by no means perfect (and you make valid points as to its flaws) but it has been selected by the industry and as such has a firm future.
There's a second, unspoken half to that sentence, which is: "unless the industry itself doesn't have a firm future".
For all its successes, it's not as though the industry doesn't have a history of expensive, mutual mistakes - mobile TV, anyone? 3G videotelephony?
There's a very fine line between the wisdom of crowds, and a consensual delusion.
Part of my job as a "professional contrarian" involves trying to draw that line.
Anonymous >> If I (as a big fixed & mobile operator - 20 million users in my case) want to migrate my network to IP everywhere: access, transport and control, what do you suggest I should do with the control layer? Whack something together with XMPP as somebody suggests above? Run my CS network until it falls apart? Just forget about services and leave them to 3rd parties altogether?
Clearly, it depends on a whole host of factors. I can't answer what's probably a $million consulting problem in a blog comment.
I don't have all the answers - and if I did, I wouldn't give them all away for free here. This blog is here for the purpose of asking tough questions - I find it generates business in helping others solve problems, or validate their existing plans.
If I just wanted to "pick a fight", there's a lot more interesting topics in the world than IMS, that I have strong opinions about.
I'd certainly question what you mean by "services". There are plenty of examples of communications services that are not based around SIP/IMS or "sessions". Wholesaling of network capacity & capabilities, in particular, or web-based functions like advertising or some way of monetising customer data.
So, to answer your question in the context of a project for you, I'd probably start by questioning the revenue forecasts and implicit assumptions you're making at the moment.
If you're interested in something a bit more concrete, feel free to contact me via info AT disruptive-analysis DOT com.
DB says: "There's a second, unspoken half to that sentence, which is: "unless the industry itself doesn't have a firm future".
And of course there's a third unspoken 'half' to that sentence - "unless the world economy itself doesn't have a firm future"
and a fourth: "unless the world itself doesn't have a firm future."
and a fifth "unless the solar system itself doesn't have a firm future"
and a sixth "unless the universe doesn't succumb to heat death"
and
and
I'm not buying it, my belief is that operators will continue to exist and sell services for a long time. They will certainly feel huge pressures from alternative OTT operators and the like but that will only serve to galvanise these behemoths into something approaching action. Will they lose significant market share? Certainly. Will they cease to exist as service providers? Poppycock.
DB: "Clearly, it depends on a whole host of factors. I can't answer what's probably a $million consulting problem in a blog comment. "
Of course, I suppose the question was only intended to point out a flaw in your thinking - there is no realistic alternative to IMS at the moment for large operators wishing to migrate control to IP.
Now, you may argue that "control" is now irrelevant but the fact remains that we have many millions of fixed and mobile subscribers and expect to serve them with IP based services in the future.
DB: "If I just wanted to "pick a fight", there's a lot more interesting topics in the world than IMS, that I have strong opinions about. "
Er? Hit a nerve?
>> expect to serve them with IP based services in the future
But you're assuming that these IP-based services are *session-based* using SIP.
Cloud computing (for example) is another category of IP-based service, yet is deliverable using various other IP access / control / QoS mechanisms. I don't see salesforce.com or Amazon Web Services clamouring for IMS.
Content delivery in big chunks is better-delivered via IP using offload to dedicated CDNs, rather than complex through-the-middle control.
(CDNs meet IMS.... a discussion for another post maybe)
DB: "If I just wanted to "pick a fight", there's a lot more interesting topics in the world than IMS, that I have strong opinions about. "
Anon>> Er? Hit a nerve?
No, just an unsubtle and shameless advertising pitch.... I pick fights on this blog partly as a way to drum up business.
Subtext: "Yes, I'm brave enough to take stand up and shots at sacred cows, unlike those other analysts & consultants huddling in fear behind the shields of consensus".
Apologies for the mixed metaphor.
DB: "But you're assuming that these IP-based services are *session-based* using SIP."
Yes of course, I really don't understand your point about Amazon, Salesforce and CDNs? IMS was never intended to host "applications" like these, however it is certainly intended that applications like these should have access to IMS applications with mechanisms like IM-SSF/Parlay-X/SOAP/SOA/JSLEE/JSR289 etc etc
Again it's horses for courses, do you have a fundamental misunderstanding of IMS' intended capabilities?
> Again it's horses for courses, do you have a fundamental misunderstanding of IMS' intended capabilities?
No, I understand the capabilities. I'm questioning the assumption that the services that can *exploit* those capabilities will really be that important.
Assuming you're the same "anonymous", I note that you've failed to make any responses to my various comments about IMS-related revenue streams. Or is it all just about cost-savings vs. legacy CS? Where's the ROI?
If I was a shareholder, I'd ask you to justify why you're investing in IMS capabilities, rather than *other* capabilities, with my money.
Will IMS assist you in attaining successful future business models? Or is it more about mitigating the decline of old ones?
DB: "No, I understand the capabilities. I'm questioning the assumption that the services that can *exploit* those capabilities will really be that important."
Communication services won't really be important in future? seriously?
DB: "Assuming you're the same "anonymous", I note that you've failed to make any responses to my various comments about IMS-related revenue streams. Or is it all just about cost-savings vs. legacy CS? Where's the ROI? "
I apologise that I have to remain anonymous, it is a very interesting discussion. We have had a valid business case for cost-saving over CS in the fixed network for some time and now have the same for FMC services. The mobile IMS business case is based on some revenue generating VAS on a small scale initially but now we are turning attention to the longer term case for IMS control.
>> Communication services won't really be important in future? seriously?
No. Communications services *that can only be provided through IMS capabilities*. Clearly, ordinary voice calls and SMS will remain important, for example.
>> The mobile IMS business case is based on some revenue generating VAS on a small scale initially but now we are turning attention to the longer term case for IMS control.
Ah-hah. So the answer is Mobile IMS Business Case = Protect voice & SMS profitability as much as possible, plus provide a platform for some other services. But we're not sure exactly what.
There are two major risks here:
1) IMS actually makes voice & SMS *more* prone to erosion, for example if RCS (aka SMS2.0) provides a worse user experience than alternative messaging services, and everyone starts using Facebook or Twitter or Skype messaging instead.
2) IMS turns out to be a sub-optimal platform for these other VAS, to the extent that nobody spends much money on them. For example, does it have built-in business model inflexibility - perhaps confining users to a "subscription" model which seems inappropriate for their needs.
The alternative is to separate the two:
1) A strategy for protecting voice+SMS profitability (eg keep and extend CS with VoLGA or CS Fallback or NSN Fast-track) plus spending more on lobbying regulators
2) Look for the best platform (ecosystem, actually) for "cool new stuff". I'd suggest something that's easily accessible to 3rd-party developers & which is optimised for encouraging innovation around devices and user-experience, rather than QoS.
I understand the need for occasional anonymity, btw. In consulting engagements I typically work under a strict NDA.
Definitely true, great post and comments!
Google Voice or Ribbit are offering much more than deployed IMS platforms, which are playing mostly as hypertrophic sip server...
Operators should shake up but don't have courage or ideas, pressure at regulatory side from internet companies in order to open the current model or to free the customers from SIMs (maybe the massive availability of wifi-enable handsets could be a first step).
DB: "There are two major risks here:
1) IMS actually makes voice & SMS *more* prone to erosion, for example if RCS (aka SMS2.0) provides a worse user experience than alternative messaging services, and everyone starts using Facebook or Twitter or Skype messaging instead."
If RCS can only offer a worse user experience it will not be launched by any operator, you can be sure of that.
So, how do I contact my Twitter friends through my Facebook account or do I need to maintain both? Will Skype provide a unified interface to Google and MSN or vice versa? I know about fring but it's interface could at best be charitably described as "OK." Perhaps we will see one grand unified client that overcomes all the UI and battery life problems you assume RCS can never solve.
DB: "2) IMS turns out to be a sub-optimal platform for these other VAS, to the extent that nobody spends much money on them. For example, does it have built-in business model inflexibility - perhaps confining users to a "subscription" model which seems inappropriate for their needs."
There is always the risk that nobody uses your latest whizz-bang service but one of the points of IMS is that it enables much more rapid service development than heretofore (no, not at internet speed but at least we can work in weeks rather than months). Your point about subscription model difficulties is irrelevant, it is the BSS that will constrain these not IMS which will allow any form of billing you can imagine/implement.
DB: "1) A strategy for protecting voice+SMS profitability (eg keep and extend CS with VoLGA or CS Fallback or NSN Fast-track) plus spending more on lobbying regulators"
I don't see this as mutually exclusive to deploying IMS - obviously it makes a lot of sense to sweat whatever assets you have to the fullest extent possible. Howevere there will come a point when a full technology refresh will happen. For a fixed and mobile operator who will very likely already have a large fixed IMS deployment in place already IMS is the logical choice. Mobile only? Well, it's a much more difficult decision but perhaps the rest of the industry will carry them along.
DB: "2) Look for the best platform (ecosystem, actually) for "cool new stuff". I'd suggest something that's easily accessible to 3rd-party developers & which is optimised for encouraging innovation around devices and user-experience, rather than QoS."
You seem to assume that IMS will preclude or stifle such innovation. My belief is the opposite. IMS provides a solid basis for such development and with the addition of technologies like JSLEE/JSR289/Parlay-X it is straightforward to expose IMS capabilities in a developer friendly way.
Ale: "Google Voice or Ribbit are offering much more than deployed IMS platforms, which are playing mostly as hypertrophic sip server..."
There are plenty of FMC implementations based on IMS that offer far more sophisticated services than GV or Ribbit. That is not to say that GV or Ribbit are not interesting and highly competitive services themselves - they certainly are. IMS is still in early implementation and telcos are ponderously slow at getting products out the door and poor at marketing complex services but they are beginning to appear.
It's true, there're plenty of FMC vendors, try to ask some company connected to those services and you'll see the customer satisfaction: one proof is that operators are selling the new platforms on price competition not on higher value...
Probably I'm out of reality but I'm not seeing all those sophisticated services... At the beginning all vendors started with the value at the core, after that you had value positioned at enablers (see OMA), and without any service coming at the horizon the vendors moved to web 2.0 integration, that means you don't need IMS but you need internet... IMS has been followed the same path of ATM: too complex, too slow, too expensive but, I must admit it, very cool and elegant (the first releases were managed by French people). Something else will appear, I don't know what, unfortunately for my pocket ;-)
The problem with IMS is that it is deemed as something new and the stupid acronym and name make no sense as to what it really is.
IMS is the continuation of the evolution to all-IP networks. But it is a very complex 3GPP way of doing things and the reason why it hasn't really been taken very far. In reality as Rel4+ circuit core networks evolve to IP and full VoIP, they would need to interconnect and facilitate billing of IP based voice (and/or video) services.
If IMS was thought of as an evolution, rather than a vendor flog, we probably would have had a lot more acceptance - albeit at a lot less fanfare about it all. But instead of following IETF ways, 3GPP and the vendors decided to overthink and complicate things so much that it required a new name and needed to be flogged at exuberant prices.
In my mind this is what functionally IMS needs to do to become a reality:
- become the actual "circuit core", i.e. route all signalling through CSCFs
- provide IP based voice interconnect between PLMNs as well as to corporate IP based PBX
- provide QoS control through a PCRF/GGSN comms - if the bearer is data based
That's all IMS should be doing.
Applications, clients, other on-top stuff that's related to IMS is not really IMS - that's the dream of the next "killer app", i.e. SMS. But if you tell operators IMS is just a replacement of your core, nobody would buy it as they already have a core and do not appreciate what all-IP really means. So that's why vendors piggy-backed all the promises and some totally stupid applications to try to sell IMS, but nobody can understand why you need to reinvent the wheel when Skype exists already. Things got even worse when FMC got mentioned.
very interesting blog posts.. i would like to believe IMS/EPC is the future, otherwise why would big telco companies bother spending millions towards the successful deployment of IMS.
Further more why is that intensive research is still currently going on toward providing hosted services (SaaS) with extended NaaS features that telocom companies can provide (e.g. presence and BOBO to pay 3rd party ASP).
Dean Bubley:There are certain control aspects for which specific IMS elements might make sense (eg policy)
This is what OMA PEEM is based on, and it's currently being integrated with IMS/EPC to provide new-value added service on demand for cloud based services.
Triggered by Dean's 1000th blog I read this blog again and all the comments.
It is funny (or disappointing) that 18 months later the post and the comments could be written today. We are not much further we 'commercial' deployments of IMS 9I mean actual money making applications on IMS). VoLTE exists largely on paper, trials have still to start. RCS is replaced by RCS-e; finally we get rid of presence. A comment suggested RCS is about the user interface and experience - well RCS client developers have their work cut out. i suggest spend big bugs and get some expectional UI designers in-house.
At MWC 2011 should be a wake up call: we had App Planet (hall 6) and Andriod infested Hall 8 (Android booth was a showcase for Apps). We have currently 100.000s of applications in ll sorts of App Stores, markets and malls; downloadbale to your device. There are millions of web services reachable from your 3G enabled device. So, who cares about the 10s of applications you can built with IMS.
Yes, some voice (or broader communication) services will be delivered to you. For the same reason as someone mentioned ' millions of dollars spent' many operators will attempt VoLTE (" I can't be wrong having spend some much. And other do it too, so it must be right").
Time has come to start using the parts and pieces of IMS that make sense and throw out the rest.
Post a Comment