Challenges for Asian Telcos: Convergence, Interoperability

( 1 Sep 2008 )

NetEvents 2008 Asia-Pacific Press Summit


During the recent NetEvents 2008 Asia-Pacific Press Summit, Michael Howard, Principal Analyst and Co-Founder of Infonetics Research, and panelists Victor Zhang, Co-Chair, Asia-Pacific, Metropolitan Ethernet Forum (MEF); Brad Booth, Chairman of the Board, Ethernet Alliance; Peter Lunk, Director, Service Provider Marketing, Extreme Networks; Sanjay Bhatia, Senior Product Line Manager, Genband Inc.; Patrick Fitzgerald, SVP Sales & Marketing, AppTrigger; and Kristin Harris, GMI Technical Coordinator, MultiService Forum, discussed in a conference debate the current challenges for telecom operators in Asia, including convergence, interoperability, and rollout issues of next-generation networks (NGNs). Excerpts:

Michael Howard: The challenges for operators in Asia are similar to the challenges for operators around the world. Telecom is in transition, maybe telecom will always be in transition, but definitely some major changes are underway right now. But I want to point out that the changes that we'll be discussing in this panel are under a capped capex world, that is, there is no unlimited funding for all of the changes going on in telecom now.

The winners are the telecom providers and the equipment manufacturers. However, these winners have to innovate, have to keep adding new services and capacity, amid growing bandwidth, while at the same time, reducing the operation's expenses and do this all within a restricted budget.

If we look at Asia capex or worldwide capex, we believe telecom spending will come to a plateau in 2009 and 2010, with the spending of that capex continuing to shift substantially toward IP NGN.

In 2008, there is a jump in Asia's telecom capex, sparked mainly by China and India spending. Part of what is driving the growth of use of more bandwidth, of course are more mobile subscribers.

And we expect mobile subscribers here growing to about 4.4 billion in 2011. There are over 400 million broadband subscribers in 2011, but that will be dwarfed by mobile subscribers. Another measure that people keep talking about driving bandwidth is IPTV. IPTV subscribers are at most 20 percent of broadband subscribers in 2011, but there is a lot of video being used without IPTV. One trend is for providers to become vertically integrated, wireless and wireline. That is making the networks become more complex.

Meanwhile, for IP NGN transformation, almost every major service providers, incumbent and competitive, are moving to packet. All these networks support voice, voice-over IP, IPTV, mobile backhaul; they also support complex services and are adding more packet, more Ethernet.

One example is that of the carrier Swisscom. From their website and many presentations, they are in the middle of IP NGN transformation. What this means, simply, is they're moving from having many single purpose networks like telephone, mobile, broadband, data as separate networks onto a single IP packet network, where all the traffic, all the applications run on one network. But of course, today, they have their NGN network, but they still have telephony, they still have mobile networks running separately. They're intermixed.

For my conclusions, there are many transitions in telecom, but all the changes I just mentioned are in a capped limited capex world. And all the winners have to innovate. And it's the challenge for the telecommunications provider to add all these new services, to keep up with bandwidth growth and be competitive in a capex world. And it's a challenge for manufacturers at the same time to produce equipment that will fit into that world. As for interoperability, how does that fit in to this picture? All this convergence in many different directions are mainly of wireline and wireless. How does your company or your organizations fill a role or help this process of convergence, or the interoperability?

Peter Lunk: I think the fundamental point you made about capex being capped and at the same time we've got bandwidth growing, is solving that problem. How we can serve more customers, and give them more bandwidth with a flat capex expenditure. I think where Extreme comes in is we're a provider of Ethernet equipment. Ethernet as a technology seems to be the best solution available today for how you get more bandwidth out to those subscribers while keeping costs down. So that's really the mantra for Ethernet and Extreme's position in the service provider market.

Howard: Great, thank you. We have here the Ethernet Alliance, the Metropolitan Ethernet Forum and the MultiServices Forum. So one question, the same question, how do you contribute to this scene?

Brad Booth: The Ethernet Alliance actually is directly focused on Ethernet technology, the stuff that comes out of the IEEE standards body. Our main objective is to help look forward in technology what people want in the next generation. We're talking about next generation now as the transfer from legacy to IP, but what we've perceived as next generation is 40Gb and 100Gb Ethernet. What we try to do is work directly with what goes on in the IEEE standards body and with our member companies to help prove some interoperability prior that equipment hitting the market.

Howard: That's necessary in this world; you're helping many manufacturers produce equipment so that it will be easier for the service providers then to buy equipment—they know what they'll be getting.

Sanjay Bhatia: Different countries in the Asia-Pacific and globally have lots of challenges on their hands. They have disparate network out there, there're the green initiatives going on, and the challenges on convergence—whether to converge networks, how to converge networks, how to do it with capex and OpEx in mind. What Genband does is we provide media gateways that enable some of these different technologies to come together. So if you have TDM, IP networks, ATM networks, which are very prevalent out there; and, as you know, PSTN is not going away anywhere soon; and then there are mobile networks that have ATM in their infrastructure; what Genband does is enable those networks to come together very effectively. In the access networks and tandem networks, when you're connecting into PSTN, the TDM-based PSTN, the ATM-based 3G mobile core, all those technologies come together. Genband is right in the middle of enabling those transitioning—and the transition to IP networks as well.

Victor Zhang: The Metro Ethernet Forum focuses on standardization of Ethernet services standards, and promotion of Ethernet services to service providers and end users. MEF is technology agnostic, which means we care about how Ethernet services are delivered to the end users, but we do not really care how it is being realised—it can be done over IP MCoAs, it can be done over Tier 2 pure Ethernet switches or a transmission. MEF does not dictate what technology is going to be used for realizing these services. Our role is more into making it easier for service providers to deploy their Ethernet services, making it easier for end users to understand what services are available and also making it easier for service providers to interoperate, to create and to send Ethernet services outside the metro area, outside of the national area, into international connectivities.

Howard: I spoke at the MEF Forum in Atlanta recently, and I chaired a service provider panel. The service providers were thanking the MEF for providing a definition, a common definition of Ethernet service, so that even inside of that service provider organization they could speak the same language. They all know what they are talking about when they say ELINE or ELAN service.

Zhang: I just want to add on to that. Today we have 146 members. What we did notice is that since last year, service providers are making a very large proportion of the new members because they're starting to see the value of what Ethernet Forum provides to them as a platform for exchanging experiences to basically learn from each other.

Patrick Fitzgerald: We're a little bit unique on this panel. We're a telecom software company squarely focused very much in line with what you were talking about in respect to convergence and the change going on in the service provider network. Fundamentally, AppTrigger has created the software solution that allows service providers to deploy, manage, migrate and then do application match ups with old to new to create something new for the marketplace. So we're focused on the convergence, very excited about the AsiaPac region relative to what's going on between wireline and wireless and the NGN build out. And ostensibly we've created a software solution to help optimize, or more importantly fix, the fundamentally flawed and broken deployment model today in the service provider space relative to services.

One of the ironies, I think, in the marketplace as we talk about is that today it takes a service provider anywhere between 18 to 24 months to deploy a service. If you think about where the next generation revenue is going to come from, the problem is it takes 18 to 24 months to be able to realize the benefit of that additional ARPU. So AppTrigger's software-based solution will focus on solving that problem to help optimize, and then more importantly help to be able to drive that additional revenue for the service provider.

So it's a very fascinating discussion when you look at an overlay player comparatively to the service provider, in the sense that they don't have the infrastructure and they don't have the legacy and the history to manage and deal with. What AppTrigger has done is create the software to really try to create a more even playing field, so that in turn they can shorten time to deployment and be able to drive the additional ARPU.

Howard: I could have touched on many things, but what has happened over the last few years for service providers is that where you add value is not in the network, but where you specifically add new value to attract customers is in the applications. Kristin Harris, the GMI Technical Coordinator of the MultiService Forum, will tell us about GMI, in particular, the "I"—interoperability.

Kristin Harris: The MultiService Forum provides a key benefit to service providers because within an IMS and NGN environment we have a lot of standards. We have so many standards you can't even shake a stick at them all. Even with the new common IMS coming out between TISPAN and 3GPP, one of the biggest problems that we're seeing is that there are so many standards and so many different ways to implement the standards, that interoperability is an absolutely critical issue. So in the MultiService Forum, we take the functional architectures that are defined by 3GPP, IETF, and make that physical.

We get physical by narrowing down the different protocol choices and defining protocol profiles, which we call implementation agreements. But we're actually creating the products that service providers want and vendors can create and guarantee interoperability between those products, which I think is a very critical issue as we look at NGN and NGN rollouts. So on NGNI, as you pointed out, since the value is moving up the stack, we're actually focusing on delivering services independent of those technologies and independent of the core architecture itself. So that's how we provide the value to the service providers in the industry.

Howard: Great, thank you. I think I would like to open the floor for questions. And yes we have a question here.

Camille Mendler, Yankee Group: When I think about current challenges for telecom operators in Asia, I think one that we haven't mentioned yet is the role of regulation in stimulating or impeding some of these wonderful technological developments that we've been talking about. I know that this panel is a bunch of propeller heads. Nevertheless, it would be great to get some commentary on some issues I submit are very important such as unified licensing regimes, infrastructure sharing and a key issue in a few countries right now, the issue of functional or structural separation. And sitting as I do in London in the U.K. and Europe, I find it interesting that many service providers think, from a regulatory perspective, that Europe provides a good template for some things that should be happening in Asia, which I might disagree with. Some commentary on that would be great.

With the example of functional or structural separation, if that happens in certain telcos, asis being discussed in New Zealand, as is being discussed in Australia, for example, and elsewhere, will there be an investment—will that impede investment in certain types of infrastructure. For example there's a big discussion now about fibre to the node, fibre to the home that's happening in various different countries. I submit that the regulatory environment may actually impede certain investments. Now if you don't have unified licensing, will you really be able to achieve certain aspects of fixed and mobile convergence. Will it be economically viable for you to do so? If you can't have shared infrastructure, as we see some shared infrastructure models in India emerging, will there be enough money out there to really invest in the right types of next generation infrastructure?

Bhatia: I think there are unique challenges and objectives in different countries, especially in the Asia Pacific region. There are the great big growth countries of China and India, which are growing at 15 million subscribers plus a month and increasing the digital divide and teledensity. They have a unique set of challenges that they face compared to some of the other countries. And so then the question becomes how do you accomplish something across different political boundaries? And it's a very challenging environment and it becomes difficult.

Regulatory environment certainly helps that. And actually in Europe, as you mentioned, there are certain regulatory environments that are actually helping some technologies take hold, one being femtocell. Just to answer your question, I think that's a very, very prudent note that you made. And it is a difficult situation across countries. But I think, by and large, the individual countries themselves are faced with their own challenges and goals. And that is actually leading to some better technologies being introduced into those markets.

Fitzgerald: I would add to that. The question in itself is probably a four day session to talk about— regulatory—but I would define that as a chicken and egg dilemma. And that's what I think the global market is facing, i.e. if I do the infrastructure investment, but then I don't get to own that and I have to share that, why should I do the infrastructure investment? And it goes round and round.

I think that one of the fascinating things about the Asia-Pacific though is that it's such a broad market, I have a tendency to talk about everything holistically as Asia-Pacific, but the reality of it is Southeast Asia is quite a bit different than Northeast Asia. If you go to Hong Kong and you sit down with PCCW and you talk to them about the regulatory environment and are they required to duplicate services over the IP network, in Hong Kong the answer is no. But if you go into other markets the answer is yes, I have to provide the same PSTN service as I do over to the IP side. So Asia Pacific has very different environments. And I guess Sanjay's point is really relative to what the goals of that for the service provider are. But I would summarize the whole thing as the chicken and egg issue.

Albert Park, Telecom Korea: In Korea, there are lot of operators who are using different kinds of models. Most of the operators are worried about one thing, and that is even though they're using this NGN system, they don't have a detailed profit model which relates to the NGN usage. So could you just explain about the profit model please?

Fitzgerald: I'm going to say something that's a little bit self serving, and it's also self serving for every vendor who's in the business: that is, our job is to go out and sell new technology. So we go out, and if you look at the historical picture of telecom over the last 20 years, there's always the next great thing. As Michael pointed out, the truth is that the network and the technology are always going to be changing. And the question at the end of the day is how you do make money out of that.

I would argue today for service providers the bulk of the revenue still resides on the voice network. And so, from an AppTrigger perspective, we take a very pragmatic view. In turn it's not necessarily about building out the next great network, it's how you leverage the assets that you currently have and augment that with new technologies to create something new. And that leads then to create the business plans that actually can pass business justification so that you can deploy the new services.

I would argue that, if you look at the slowdown of IMS build outs, the reason that IMS slowed down is because the actual business plan to build out the network infrastructure never made sense, because then it would duplicate all the revenue applications over in the IP side. And at the end of the day, there is no business justification to go create brand new applications, when you already had those applications and you are just going to create an IP environment and not necessarily get the ARPU.

So, I think having a pragmatic approach on it as opposed to a holistic and architectural is a much better way, from a business perspective, for telecom companies to be successful.

Booth: I think one of the other things we are seeing now is, with the transition naturally toward an IP network, a lot of the business models that you can use for getting your money from your voice network are changing. People are starting to use other applications that are IP based and eliminating some of that revenue stream. So as you look to move forward, you seriously have to consider that if people are moving to IP based networks and IP based protocols and applications, are you still going to be able to maintain that same revenue stream that's been supporting that build up? You have to know that there's a time where that revenue stream is going to taper down and you better have that next structure, that next network sitting there ready to figure out how to build and get revenue off of it.

Howard: Yes, that is part of the answer. Fixed line operators around the world are all seeing their voice network revenue going down already. They're forced to make a move to do something. So part of the rationale is that they have to move to a single communications network, everything is packets, so it's natural to go to the IP NGN network. They must do that to survive. Now they must have new answers for where does the new revenue come from to make up for the declining voice revenue.

Harris: I would also argue that IMS is not seeing a slowdown in rollout. China Mobile has just joined the MultiService Forum to host GMI to actually test their IMS and end-to-end QOS over their IMS architecture. They're planning a big rollout. And I think that IMS inherently offers an excellent opportunity to produce new services and not have to change the architecture every single time you want to put in a new technology, new service.

Howard: It's probably the case that IMS did not taken off as fast as people thought. There's a slowdown in expectation.

Bhatia: And maybe I can add—I share Patrick's views on this. I think as an equipment manufacturer ourselves, our job is to show value for what we provide and why transition to next-gen networks. And I think it's really the ultimate proposition in the environment when capex, opex controls are in check. It's our job, and we do this, to show why moving to an NGN, whether it's the IMS or other conversion networks and other aspects, does make sense. And then we do that.

Zhang: I guess the deep penetration of broadband services has been a double-edged sword for service providers around the world. What is happening is previously service providers were right in the middle of the value chain. But now with the penetration of broadband on the residential side, there is the competition coming from all the top service providers.

On the business side, with migration towards Ethernet flat or more transparent technologies, Sis are really coming into the space to try to take a share of the market from the service providers. So far, it's been the fixed operators, network operator, who has faced the same issue. But with the 3G opening out the bandwidth for the end users, the mobile operators will very soon see the same challenge as well. So what is important is for service providers moving to next generation network in a broader definition to able to find a way of bringing them back into the value chain and becoming a central part of that, so that they can really generate revenue streams from content providers and subscribers, just really get back into the game.

Fitzgerald: If I can add, I think the heart of your question is this marketplace discussion regarding the killer application. And then what happened was everybody realized that we don't know what that is, or if there's the opportunity to do triple-play or a quad-play. The bulk of the revenue is still being driven by voice, which is from traditional telecom. The discussion now has become, "Okay we're not going to worry about the killer application; we're going to worry about the killer environment." How do we create the killer environment so in turn you can reach some of the markets very quickly and in turn be able to create personalized services for the end users? I would argue that there is a slowdown in IMS.

I think it's apparent; every service provider I go talk to talks about that they've done trials but they haven't gone full headlong into actual deployment. And at the end of the day what they're realizing is that the network is going to be converged. It's going to be PSTN and wireline and wireless and NGN. And it's going to be a soup, the complexity of it. The issue then is how do you manage that and how do you in turn create that environment so that you can create a better business case to be more profitable. I think that's where the heart of the matter is in telecom in the next 15 years, which is how do you create that killer environment for the service provider.

Howard: We're going to wrap up the panel. I'll make one last remark and that is, to continue what's been said, in our observation of IP NGN networks, talking with service providers around the world for the last few years, the first focus for IP NGN is to save capex and have more efficient operations. That's the principal driver to head towards an IP NGN network.

However, what we're seeing recently is that some service providers are now using that single network to build an applications platform environment. So now that you have one network rather than many networks, they are creating a new environment to add applications easily and quickly.

CAPTIONS

Figure 1: Conference Debate Chair Michael Howard, Infonetics Research.

Figure 2: Panellists from left: Victor Zhang, MEF; Brad Booth, Ethernet Alliance; Peter Lunk, Extreme; Sanjay Bhartia, Genband; and Patrick Fitzgerald, AppTrigger. Not in picture, Kristin Harris, MultiService Forum.


Click here for the illustration:

Figure 1, Figure 2