IMS, I’m breaking up with you

November 27, 2013

Where is IMS headed and what should engineers be doing?

[When Chad Hart told me he had a post about IMS, I just couldnt say no. I was more than happy to publish it here. It isn’t his first post here, and hopefully not the last.]

I was sitting next to Tsahi at the WebRTC Conference and Expo in Santa Clara last week during the IMS & WebRTC sessions moderated by Victor Pascual. The discussions of Over-the-Top (aka Internet) services based on WebRTC vs. the 3GPP’s IP Multimedia Subsystem (IMS) reminded me of a blog post I wrote in September 2012. It was considered way too harsh by my colleagues at the time so unfortunately it was never published. Given the resurgence of the WebRTC plus/vs. IMS topic, and the fact the audience can handle it, I think now is a good time.


IMS, I’m breaking up with you

IMS, I was instantly enthralled with you when we first met. Remember when I wrote this in my IMS report at VDC back in 2005?

IMS is a unique catalyst and potentially disruptive technology to the telecom community. Never before have so many diverse stakeholders, including competing vendor and carrier groups, agreed upon such a large set of overarching network standards. Because of its technical robustness and overwhelming support from all parties involved, IMS is destined to be a success.

Innovative new services and applications with the reliability and quality of the telecom network – it sounded so good. Everyone across the whole telecom ecosystem was talking about how great you were – everyone!

We all thought you would usher in a new era of telecom. We all believed in you.  We really wanted you to succeed.

That all seems so long ago now.


Things did not turn out the way you said they would. Where are the new services and applications? Where is the disruption? I thought you said would be a big part of mobile services? I still see the same old fixed voice services we have had forever.

Things have really changed while I waited on you. Traditional voice telephony isn’t as cool as it used to be. Voice ARPU and MOU is declining in mature markets – people are finding other ways to communicate. A lot of those innovate services and applications you got me so excited about – enhanced phonebooks, click-to-dial, instant picture & video sharing, video telephony, intelligent contact centers, location based services, etc. – are available today, but not from you! OTT figured it out, went off and did it while you were doing who knows what. You could have made such a difference, yet you did nothing…

I know you’ll say at least we have Joyn? But are you really proud of that? Remember your first RCS attempt with all the cool features back in 2008? Here we are 4 years later and you are just starting to get around to rolling out a subset of those features!  Sadly I think you only launched that because you were jealous of OTT.

I know what you’re going to say next – “just wait for VoLTE and then everything will be fine again”.  VoLTE always seems like it’s just around the corner, but it never shows up. I’m sorry, but I can’t wait for you anymore. I’ve given you so many chances. I’m not even sure if you are even capable of changing – sometimes it seems like you are stuck in the wrong decade, maybe century!

Oh, and one more thing. I’ve found someone else – its OTT. I know you’ll snicker and ridicule OTT because it doesn’t live up to your standards like you always have, but OTT has come a long way. OTT has replicated pretty much everything you do and more, and OTT’s getting better every day. You’re right, OTT doesn’t care about your revenue-per-subscriber models or universal interoperability, but it doesn’t need to – OTT is not trying to be like you. OTT is doing all the things you said you would do without your baggage. If you would wake up you would see that even most of your telco buddies are now hanging out with OTT behind your back.

Well, I know a lot of my friends will keep hanging out with you so I’m sure I’ll still see you around. Stop by and pick up your acronyms some time. Hopefully you can bring yourself to be friends with OTT one day.



IMS breakup


What do I think now?

I wrote that 14 months ago after several months of focused work on figuring out the OTT (a term I always hated) and Telco-OTT market. I had been intrigued by how the traditional telco community underestimated the OTT model for years. Many who leveraged VoIP’s disruptive potential more than a decade ago forgot that they too could be disrupted.

Ironically, several months later my responsibilities were expanded to include IMS! More recently I have been more focused on just WebRTC, which I see as a natural evolution and maturation of OTT services.

So what do I think now? Well, first I should say I am a pragmatist. I do not expect the 10’s of thousands of engineers who have built their careers around IMS to just up and quit their jobs.  Infonetics reports there are over 200 million subscribers on an IMS core and that number has been growing steadily in single digit percentages. Those services keep getting investment because they are profitable. Those users certainly would not be happy if they were just shut off one day.

There is a place for IMS. If success is measured in billions of subscribers across every regulated telco then IMS has no hope. If success if measured in 10’s of millions of subs per operator then the prospects for IMS are not so bad, particularly if IMS can learn to accommodate OTT technologies like WebRTC.

That being said, the future is undeniably the web, which I define as anything that leverages the HTTP protocol. The web has figured out how to give anyone access to the world’s information in a fraction of a second. The web handles 100’s of millions of daily users seamlessly via a highly distributed, but extremely reliable, virtualized infrastructure. The web supports hundreds of millions of unique services – 625 million websites + several million mobile apps. Now the web is applying this vast amount of awesomeness to the communications industry that spawned it decades ago. The web will be undoubtedly be successful in real time communications – the problems of real time telephony are miniscule compared to what the web has accomplished already.

So what do you do if you are one of those IMS engineers? Go learn JavaScript. Build a website. Contribute to open source projects. Master WebRTC. Expect the same level of excellence from your own network as you do from of the web services you rely on every day. Then, if it turns out that IMS does not survive this new web-oriented age, then at least you will.

You may also like

Fixing packet loss in WebRTC

Fixing packet loss in WebRTC

Your email address will not be published. Required fields are marked

  1. Another great post, Chad!

    I never fell in love with IMS and the Telco industry lost a great opportunity to create the baselines to sustain agility and innovation. WebRTC is another opportunity we have to build a new Telco world where Data and Communications are linked naturally. We must keep standardization activities to a minimum but we should still aim to ensure fully interoperability among Service Providers. It is time to move on and think about the post-IMS era.

    1. Agreed on everything, but I am not sure WebRTC-based service interoperability among service providers should be a focus at all at this stage. When a service is successful then it earns the right (or demonstrates the demand) to federate. Alexander Graham Bell certainly did not have universal federation in mind when he started Bell. The first federation into a semi-national system only came about after AT&T bought up all its competitors or drove them out of business. Federation among the various PTT’s took longer, long after there was demand, not before it.

      WebRTC’s media engine certainly does have applicability to be used as a replacement engine in today’s interconnect networks for existing voice services, but I am not aware of anyone working on that.

  2. As one of the panelists in the WebRTC Expo session mentioned at the beginning, it’s gratifying to hear this kind of reaction from the audience. The panel seemed sharply divided, with Evan and myself representing those not waiting for IMS, while ALU and NSN represented the “just wait! it’ll be great!” crowd. As application developers, my position is that we’ll use whatever technology we have available to solve problems for our customers. IMS today, for all its potential, is just wireless PSTN. Until that changes (and carriers open up to allow outside applications), we’ll keep moving forward with things like WebRTC and OTT options.

  3. Dear Chad

    Thanks for your letter. It would be too easy for you to send me something full of bitter recriminations and name calling, so I really appreciate you being so constructive. I always hate breaking up from a long term relationship, but sometimes these things happen. I’m sure you know there are others in love with me, so it is not surprising that you might feel a bit betrayed.

    Looking back on the trail of partners and lovers I have had, I do think I need to put up a a little defense. When I was young and raw, when I had vision, then as you note, everyone loved me. I was the burning hope of simplicity to cut through a complex new world. But the more I was used and abused by my cortisans, the more I became scarred and twisted by their various attentions. It got complicated, messy and confusing. I found some shelter with fixed networks, but there I was not anything new and I was hidden away, unable to really make a difference.

    Mobile is where I have been confused most. Sometimes I think that it is me that has got it all wrong, but when I look in the mirror in those darl times I see that no, I am the answer to mobile’s questions, but it is mobile that can not bring itself to adopt me. It is not that I am the wrong answer, it is that they don’t want to address the question properly. A relationship never works when the two parties involved don’t communicate, which is ironic when communication is what I am all about. Mobile has made me promise after promise, they tell me ‘tomorrow is your moment’, but tomorrow never seems to come.

    VoLTE keeps me hoping. It remains the one thing that I think might let me burst into the lime light, and whilst I understand why you have had your head turned by a younger, lithe, flexible technology, I do have some warnings.

    Remember why you fell for me in the first place Chad. Remember all the promises I gave you about interoperability, about control, about being able to manage consistency because I was finite and understandable. I might have got a bit more complicated, but I still have keep to those promises. WebRTC might deliver you amazing new experiences, but it is wild and if you thought my challenges were hard, try and make a WebRTC call that is interoperable when you don’t know what the other end of the call is. If you want to be able to do that, then you will need to reign WebRTC in a bit, place some restrictions on it, make it understand the places where I went wrong and learn from the mistakes we made together. Because WebRTC claims to be simple because there is nothing between its end points is what will make it dangerous complicated to fathom when you look to make it work for more than one off use cases.

    The problem is, when I look at WebRTC I see all the same mistakes we made but with ten times the number of variables. I had control protocols defined, I had understood capabilities, I even knew how to agree codecs. WebRTC has none of that. That is exciting to play with, but once you are married to it, the reality of who you are living with might kick in. I may be wrong, but I think you may find it is already so out of control that it is only going to be tameable in scenarios where you can control both ends of any session it is applied to.

    So I know it is over for us. It’s not you, it’s me. But please take care with your new love. It may be exciting and fresh, but that comes with unpredictability and volatility too. As it matures it may settle down to be so very rewarding and you may go on to have a beautiful ecosystem together. However, if WebRTC turns on you, keep in mind that I might be able to help. I have principles that WebRTC doesn’t, which are important to a man like you. An open relationship is not out of the question.

    I’d even consider a threesome.

    Yours forever


    P.S. I’ll pop buy and collect the acronyms next Tuesday. You can keep the Billy Joel CD though. You always liked it more than I did.

    1. Nice, mature response IMS. I hope we can be amicable since we will still run into each other all the time.

      But what is this talk of Open relationship”, “Threesome”? You know my basal instincts to be free and open find that appealing. But then again I remember you have said that before. Remember when we did some OneAPI together to spice things up in our relationship by bringing in some new developers? We all remember how embarrassing that was…

  4. “That being said, the future is undeniably the web, which I define as anything that leverages the HTTP protocol.”

    To expand or detour on that thought, the future is undeniably the cloud, meaning with an IP connection I get the services I want to any device over any broadband network. This is not just a reality for consumers and businesses, but service providers as well. Building and running a network to offer next gen voice is not the only option.

    The cloud can be the source of that technology; the service provider can outsource and focus on their business (and spend that capex on bigger, better broadband). Why build the 3rd/4th/5th generation voice network in the face of margin pressure and prevailing winds of change?

    IMS – big telco, big capex
    Cloud voice platform – every other provider
    Telco OTT – extension of either of those solutions to any device, browser

  5. To whomever wrote the IMS letter: kudos. Both of these are wonderfully (and hilariously) written and touch on the exact points we spoke about during the panel at WebRTC Expo. IMS has value in its current deployments and shouldn’t be written off, but WebRTC/OTT and the emerging ecosystem take a lot of wind out of the sails of future deployments and RCS. The IMS community refuses to see this, as someone on the panel mentioned that IMS could accomplish much of the WebRTC potential with a browser plugin. I’m still kicking myself for not highlighting that!

    I’d like to hang for a minute on the issue of interop being a barrier to adoption; I just don’t find this to be a strong argument. Most current WebRTC deployments are single-site, but that can be attributed to the standard still evolving and people developing/experimenting constantly. A de facto standard (SDP, etc…) will emerge from the strongest players (as AWS has done for cloud services) or from the ecosystem itself.

    The common use-case example in IMS of a doctor requiring interop ubiquity to ensure he can reach his patients is a bit ridiculous – even right now with the normal PSTN, I can rarely reach a human at the hospital. Nor does my doctor *want* to hear from me; she’s the constrained resource, not me. But let’s say we want to integrate IBM’s Watson into a WebRTC app for initial triage, but can escalate to a doctor in the event the automated diagnosis is grim? And if not then simply leaves a flag asking the doctor to respond at their convenience? Now imagine implementing that in IMS.

{"email":"Email address invalid","url":"Website address invalid","required":"Required field missing"}