IMS & Next-Gen Networks: The Year in Review and What's Ahead

In the world of next-generation networks, IP Multimedia Subsystem is the 800 pound gorilla. IMS was designed as a common service architecture for both wireline and wireless services. Based on such protocols and technologies as SIP, SDP and DIAMETER, it enables the quick and simple production and deployment of news services with the aid of service building blocks, and offers coordinated and ensured Quality of Service (QoS) and correlated accounting/charging among the service, session and transport layers. It also supports common interfaces to application servers for accounting/charging, security, subscriber data.

 

Although nearly all carriers world-wide look upon IMS as an inevitability, interoperability issues and infrastructure costs in some countries have made the 800-pound gorilla appear more like an 800-pound ground sloth.

 

As Tekelec’s Vince Lesch, Vice President, Product Marketing, says, “Operators are deploying IMS much slower than industry leaders originally anticipated. Industry leaders have realized that the IMS business case doesn’t work because it requires too many components to be deployed. Given that full-blown IMS deployments are still years away, Tekelec is currently working with a number of service providers – who are still investing heavily in their next-gen networks – to make these networks more effective and efficient. As a result, operators are trying to get the benefits of IMS without all the costs; operators have no plans to rip out their existing networks, but rather are deploying SIP Signaling Router [SSR] technology in order to leverage their existing investments for as long as technologically and economically feasible. This also lays the groundwork for operators to realize the benefits of IMS today and pave the way for deployment of all of the components of an IMS network: CSCF [Call Session Control Function], SIP routers, and so forth. In fact, Tekelec has seen an uptick of interest in its SIP Signaling Router over the past year, which signifies the intention of service providers to make the transition to next-gen networks.”

 

“In addition, as you likely know, Tekelec has participated in the past five IMS Plugfests – with its TekCore having been the reference CSCF for all five – and will be participating in the one taking place January 12-16, 2009 at the University of New Hampshire InterOperability Lab [UNH IOL] in Durham, New Hampshire,” says Lesch. “The Plugfests drive innovation and interoperability in IMS, and creates an open forum for vendors to come together to validate their products and ensure that they’re developing industry-leading products that will help smoothen the transition to next-gen networks.”

 

Tekelec also participates in the SIP Forum SIPits events, as well as the MultiService Forum (MSF) GMI events. These events are additional venues that allow vendors to bring their SIP and IMS products and services to an open forum in order to conduct interoperability testing, with the goal of eventually making these global standards. Lesch says that Tekelec’s CSCF will be included in the upcoming MSF GMI 2009 event.

 

Manna from Mobile Mania

 

NextPoint Networks was created in 2007 via the merger of Reef Point Systems and NexTone Communications. In September 2008 it was acquired by Texas-based GenBand. What they got was NextPoint’s Session Border Controller (SBC) technology that provides security and interoperability at the edge of service provider networks dealing in IP-based real-time services such as voice, video, and multimedia over wireline, cable, or wireless networks. It supports network architectures including NGN, PacketCable, IMS, and FMC.

 

Natasha Tamaskar, Vice President of Product Management for the NextPoint Mobile and FMC solutions says, “From our perspective, NextPoint was a company that came together from different backgrounds, a mobile one from the ReefPoint side – where I come from – and an SBC one from the NexTone side. We have all been watching the IMS play for a while and from our perspective IMS and the transformation of the world’s network infrastructure is not something that’s happening overnight. It’s a gradual adaptation of new services and new technologies. Initially, when IMS got rolled out, it was really more of an architectural adventure. People said it was a beautiful architecture and thought it would solve all kinds of problems. But there really wasn’t enough discussion to work out the business case for it and how it’s going to solve other people’s problems, and why would people start adopting this new technology anyway? Since 2007 people have spent a lot of time on the particular problem of why would anyone want to deploy IMS. What are those killer applications that would encourage operators to move to IMS.”

 

“My focus has been on the mobile side,” says Tamaskar. “In the mobile market I see it’s all about determining what the mobile user is looking for. What is the nature of the mobile experience? More and more, users want all kinds of ‘cool’ applications on any device at any location, at any time. If you look at it from a carrier’s perspective, if they need to deliver these applications, what is the easiest mechanism to do so? The answer is IP. This trend existed even before IMS, of course. First you need to have an infrastructure which is IP-based so you’re able to claim you can deploy IMS. So carrier cores and active networks are moving toward IP, and although it’s a few years away, the big push toward high bandwidth LTE [Long Term Evolution] is coming. There are also large bandwidth-hungry applications under development and deployment. All of this will help drive the adoption of the all-IMS core, but it will be a gradual process.”

 

“Interestingly, we’re starting to see that applications not traditionally associated with IMS such IPTV are now, even from an architectural perspective, being overlaid on top of this infrastructure,” says Tamaskar. “Architectures evolve to be more application-friendly and display more suitability for things other than VoIP and we will see more of that going forward. New applications will drive advancements in the network core. Things will probably be driven more from the mobile side than the fixed side. Initially, IMS came in from the 3GPP but at the time you really didn’t have access networks that were 3G at all. Ironically, even today most access networks on the mobile side are circuit-switched. Greater adoption then occurred on the fixed side, at least in terms of more VoIP-type deployments, which have more potential for moving to an IMS-type environment. But everyone says that it’s once again the mobile devices driving things along with the users who want to run mobile applications such as video share, photo downloads, IM, and so forth. So once again IP becomes more interesting on the mobile side, not just for pure voice applications but these multimedia apps too. If you can drive that through common code and provide a common infrastructure for billing, that’s where the actual value-add enters the picture.”

 

Asian Adoration

 

“We are working both in Asian and European markets as well as U.S. markets,” says Tamaskar. “In the U.S., our big Tier-1 carrier customers are starting to progress from the lab/testing stage of IMS to talking about actual deployments in 2009. The Verizons and AT&Ts of the world are traveling that route. People are suddenly taking all of this more seriously. IMS and the transformation of the network are not going as fast as was initially perceived, but it certainly is coming along. In the Asian markets the situation is very different. People there have no choice but to roll out IMS cores because they see many threats from Skype video and voice. They see that people are looking for a lot of multiservice offerings, whether it’s for fixed, mobile, cable or IPTV type solutions. And they want to maintain subscriber loyalty. The demand is greater in Asia, and it forces the operators to roll out more services. Interestingly enough, we’ve seen people using different ways of deploying IMS. Some say that they’ll take the first application that people like and run that end-to-end in IMS-friendly form. So no one is really initially designing a complete core for everything and then determining what applications can be rolled out. It generally starts with a silo scenario, and then it broadens to become a more flattened architecture, but things are rolled out on an application-by-application basis, not all at once. Current users are sometimes kept on a circuit-switched path, but new users connecting to the mobile core are rolled into an IMS scenario.”

 

Picking and Choosing

 

Starent Networks provides networking solutions used by mobile operators to furnish multimedia services to their customers, such as video downloads or streaming, Internet access, Voice-over-IP, email, mobile TV, video and photo sharing, gaming, and others.

 

Andy Capener, Starent’s Director of Marketing, says, “Obviously, we have a mobile view of things such as IMS, which is a unifying service architecture for fixed and mobile communications. We’re an eight-year old company that focuses on mobile multimedia core network. We sit between the radio network and the IP services. We provide the core elements that the operator requires for their packet network. Many of our early deployments and success has involved access gateway solutions for CDMA and UMTS operators, acting as that subscriber gateway and subscriber management point into the packet network. We differentiated ourselves by improving performance in the network offering integrated service capabilities at the access gateway that can shape the subscriber transaction, as well as putting more intelligence in the core network so that the operator knows what the subscriber is doing and then can shape those individual transactions or that individual subscriber’s experience. So, we’re a mobile infrastructure provider. We have over 80 mobile operator deployments in approximately 35 different countries with customers such as Sprint, Verizon, KDDI, Vodafone, China Unicom and many Tier-1 and Tier-2 operators.”

 

John DePietro, Director of Product Management, says, “Because IMS assumes you plug into an IP-capable radio network, we fit into a very good position in the network to provide IMS core capabilities,” says DePietro. “Our focus is around providing an IMS core set of functionality. You may have seen our announcement regarding our session control manager platform which is essentially the same platform we use for our Packet Core solutions, but runs essentially the SIP infrastructure you need for an IMS core, through all of the CSCF [Call Session Control Functions]. So basically it’s a specialized SIP proxy and registrar. We can handle the entire SIP core for the IMS. We’ve offered this product for two years and we’ve done extensive interoperability testing and in fact we’re outstanding members of the IMS Forum [now the NGN Forum]. We do a great deal of interoperability testing with all kinds of vendors that participate in their Plugfests, and we participate in the Global MultiService Forum’s Global MSF Interoperability [GMI] events, which every two years also tackles SIP interoperability. So, we’re very active in the promotion and development of SIP as it pertains to IMS core multimedia applications.”

 

“Our focus really is to provide various operators the ability to implement IMS on the same platform in a very straightforward way,” says DePietro, “with many features that they love about the core products, which is really scalability and intelligence. That’s our strategy.”

 

“The big challenge in the mobile space for IMS is really the separation of CDMA and UMTS,” says DePietro. “Basically it comes down to applications and clients, and of course the business case for each operator. The important thing is that IMS is really a universal service architecture for any operator wanting to deploy multimedia applications in an open, standard way – which all operators want to do – and to use that in the way that they need to use it in terms of the applications that they deploy. Everybody thought that if you just went ahead and built IMS, all kinds of applications would soon appear and there would suddenly be a great new world of services for users. The reality is that every operator uses IMS based on the applications they’re attempting to deploy. They’re looking to use IMS for applications where the benefits of IMS are most advantageous. So they’re really viewing IMS in terms of the questions, ‘What services do I want to deploy?’ and ‘Are those services best served by IMS networks?’.”

 

“We all know that, over time, operators will have to adopt IMS,” says DePietro. “If they’re coming off of a circuit-switched infrastructure and they move to an all-IP architecture – which they are now doing quite aggressively and which we’re right in the middle of – at some point each operator will hit a ‘tipping point’ where they can no longer run through networks. They can’t afford to be competitive and run a circuit network, a separate service architecture that silos applications, and also offer all of the IP and multimedia services. At some point the operators will have to capitulate and turn the corner, implementing a full-blown IMS for all of their traditional voice services, enhanced voice services and all of the multimedia stuff. It’s all going to happen – the question is timing, which depends on each operator’s case. Typically we find that greenfield operators, particularly operators moving into Fixed-Mobile Convergence [FMC], are actually much more aggressive in terms of their IMS deployments. In fact, we think WiMAX is a very great market for IMS because if a greenfield is deploying WiMAX and 4G LTE, there’s no other way to do voice. In that case, no operator would implement something that isn’t IMS compatible. Starent is focusing on this area. We’re well positioned in WiMAX and 4G. A bunch of greenfield operators are appearing with the cable operators, and we think that is a great opportunity for IMS.”

 

“Then again, you also have operators who have taken IMS and are using it for those applications that are best served by that architecture,” says DePietro.

 

Giving Developers a Helping Hand

 

Over at Oracle (www.oracle.com), Ty Wang, Senior Director of Product Management for Service Delivery Solutions, says, “We acquired BEA Systems in early 2008 along with what used to be called the BEA WebLogic SIP Server, used in an IMS network. In fact, we’ve just announced the 4.0 version of what we now call the Oracle Converged Application Server. In terms of milestones, what’s different about this product is that it’s the first product under the Oracle banner that actually has much of Oracle’s product capabilities. For example, we have something called ‘Oracle Coherence’ which we got from the acquisition of a company called Tangosol. It’s a data grid software product that boosts application performance by providing fast, distributed access to frequently-used data. We added Oracle Coherence into the former BEA product to increase the capabilities involving real-time data caching as it relates to performance increases, high availability and such, particularly concerning carrier networks and IMS. It’s part of our strategy of combining software code that we have within the greater Oracle organization. As you can imagine, we have lots of software from acquisitions. We can thus assemble these together and synergize them without OEM or partner agreements. We can just call in one of our engineering groups and tell them to give us the code, and soon it’s done. So we have a great ability to be responsive to marketing and customer needs.”

 

“Secondly we support JSR 289, which is the SIPServlet specification 1.1,” says Wang. “One thing we see in IMS and the whole convergence of the Internet, web and telco worlds is that there needs to be the realization of standards that actually can combine those worlds. It should be manifested in a product that’s converging web and telco ‘containers’, such as a SIP application server container, or a programming model – something needs to be the ‘forcing function’ for standards to facilitate converging these worlds. Well, the programming and execution standard for the converged web and telco containers is this standard called JSR 289. Oracle is actually the specification lead for JSR 289 and with this new release, we feel it’s the world’s first product that actually supports JSR 289. In terms of overall impact, this means that we want to ensure that there’s something truly standard to which people can refer. JSR 289 is now done and our products supports it. The standard had a lot of input from telcos, network equipment providers and also IP folks such as Oracle. JSR 289 is really starting to make it easier to build converged web and telephony applications, particularly around things such as IMS. It answers the question as to whether you can expose presence information in a very IP-like manner for multiple networks. Can you now go seamless from a push-to-talk session to a multi-party full-duplex conference call? Yes. Can you do a lot of this traversing between wireless and WiFi networks? Well, a lot of those apps will be developed and built by web developers and not necessarily by telco developers. So JSR 289 is definitely one of those standards helping in the development of applications.”

 

“In line with JSR 289 is our enhancement of service creation tools,” says Wang. “We’ve actually injected a lot of what we call the Oracle JDeveloper products into the support of our Converged Applications Server 4.0. A developer can now use very familiar development tools such as JDeveloper and the Eclipse environment to build these converged web and telco services for IMS or plain IP networks. It increases our ability to reach out to more application developers. There’s definitely a need for more app developers in the web world to build these innovative apps for both the web and telcos. This proliferation of applications development helps the IMS cause.”

 

“Finally, we’ve incorporated the JVM [Java Virtual Machine] into our release,” says Wang. “It helps to improve the latency, garbage collection and real-time effectiveness of the product, which is very much in line with the web and telco application needs.”

 

“Customers use our Converged Application Server to migrate away from SS7 legacy IN networks and put their applications onto commercial off-the-shelf hardware,” says Wang. “Some U.S. carriers want to migrate away from their tandem 4ESS switches to a more IP-based approach. Those class 4 switches are really expensive to maintain, the OpEx is high, and there is no money for CapEx. So we’re a good option for them. Also, customers are looking to build advanced hosted applications that run over IP networks. InterCall, for example, uses our server to handle over 400 million hosted conferencing minutes per month.”

 

“Oracle is even using the Converged Application Server for its own contact center applications and other communications-related applications,” says Wang. “So you can say that we’re ‘eating our own dog food’.”

 

“Finally, a European operator just bought our server to use for their IMS network,” says Wang. “We’ve already worked with companies such as KTF [Korea Telecom Freetel], several European operators, and we’re starting to work with some of the cable MSOs on IMS.”

 

Wanted: Standardized Standards

 

Comptel Corporation is a major vendor of Operations Support Software (OSS) for telecom service providers worldwide. Comptel's expertise in inventory, provisioning and activation, mediation and charging enables service providers to focus on delivering innovative services. Comptel in 2008 acquired the U.K.-based IP fulfillment specialist Axiom Systems.

 

Simo Isomaki, Comptel’s Director of Business Development, says, “Operators have been moving from trials to deployment mode with multiple services, but there’s a growing worry, as reflected in Gartner’s trend charts, that there’s still a lot of things to be done in terms of IMS and standards. For example, it’s a known fact that one of the early adopters that went into IMS big time, KPN, now complains that there are too many vendor ‘extensions’ that make the whole business case of IMS a bit too complicated. IMS doesn’t ‘fly’ so easily for them, because they can’t really integrate what they have into the kind of multi-vendor environment that IMS promised. You can’t have everybody using incompatible extensions. There must be a real standardization of the software core or integrating technology. That’s a real worry, and there’s been a great deal of interest in solving this problem. That’s good, because the only way IMS can deliver on its promise is that it must truly create an environment wherein interoperability, standardization and industrialization of the interfaces between the infrastructure components is such that everything can seamlessly work together.”

 

“Early in 2008 I was in Chicago having a discussion with a U.S. cable provider that had run through an IMS trial on the network side,” says Isomaki. “Their only worry was that they couldn’t easily make a business case for deploying the IMS infrastructure because of the integration cost. They discovered that to deploy a new service they would need multiple interfacing points to the network and most of these would be proprietary in nature and would have to be developed and refined by the software vendor, such as us, doing the service activation or doing the mediation for charging purposes. That emphasizes how we need to support a specific standard. Once that occurs, service providers can focus on actually developing and deploying services. Having said that, every operator I talk to says there’s a lot of promise and hope, and they reach for something like IMS and NGN and they’re working to go there, but today it might be difficult to really devise an extensive IMS deployment and put it in place with all of its bells and whistles and still manage to create an environment that delivers on the promise of lower cost. That’s my rough assessment of how I see IMS today.”

 

“Even so, there’s a lot of excitement surrounding IMS,” says Isomaki. “Every day I discover a new service that’s available to run in an IMS environment. Even open source projects are delivering software that references IMS core technology. There’s a lot of momentum around IMS, but at the same time, there’s a great concern that IMS will have difficulty delivering on its promise because most of the vendors in the space aren’t strictly following the standard concepts. The NGN Forum’s Plugfests are a way to certify something on the network side; they are good at ensuring that the network itself works, but when you need to deploy services onto IMS, you run into the problems with integration points and interfaces. In my discussions with the cable provider in Chicago, the CIO said that their worst-case scenario centers on the fact that they have 20 integration points that must be attended-to in order to launch a fairly simple service. So they need to touch 20 different pieces of equipment, each having a proprietary interface that deviates a bit from the standard. That means that the cost of the integrations for stepping up the services will be fairly high compared to the relative value of the service itself. He wanted to know what the industry was doing to resolve this situation. To be honest I don’t think that the OSS/BSS industry has yet been successful in dealing with this issue. The TM Forum, which develops widely-adopted management standards for telecom OSS/BSS, would probably be the best place to start the standardization process. But there hasn’t been any major activity to date in this area. Still, the Plugfests do help ensure interoperability and an increasing number of software vendors are participating in them.”

 

WiMAX and LTE – Two Sides of the Same Coin?

 

Even before IMS becomes ubiquitous in the world’s networks, new and more powerful forms of transport will appear to serve the rising tide of bandwidth-gobbling mobile devices and applications.

 

Cisco Systems, for example, offers a complete end-to-end broadband wireless solution that features best-in-class Mobile WiMAX technologies, tightly integrated with Cisco’s IP NGN transport and service delivery infrastructures.

 

Cisco’s Kittur Nagesh, Director of Marketing in the Service Provider Marketing Segment, says, “There is some confusion, some of it unwarranted in terms of the positioning of LTE [Long-Term Evolution] and WiMAX. Sometimes I see headlines such as ‘LTE Washes Out WiMAX,’ as if it were a boxing match. The context for all of this concerns what’s happening with the Mobile Internet. In the last ten years, new applications have appeared that have really driven the need for high-performance data transport. The mobile operators have gone beyond mobile voice, offering all kinds of applications such as SMS, that have become pervasive. As we go forward, the data and application-driven aspects of the mobile Internet will increase even more profoundly, and will enable richer applications, both for collaboration and general productivity on the business side, or infotainment on the consumer side. The radio networks which hitherto were largely doing mobile voice have now adapted or are in the process of adapting – depending on the technology you’re talking about – to enable these high performance applications. Applications that demand more bandwidth drives high performance Internet architectures, personalization, and new powerful radio technologies such as WiMAX and LTE.”

 

“From a Cisco perspective, these new radio technologies and even the existing 3G technologies such as HSPA or EVDO, all enable a new wave of applications which will call for an exploitation of the virtues of IP networks from an end-to-end standpoint,” says Nagesh. “That is good for Cisco; whether it’s LTE or WiMAX, it really doesn’t matter. We have enabled networks over the last ten years through a portfolio of routers, switches, gateways, service engines, application awareness, deep packet inspection, and so forth, to help bring forth these capabilities with the notion of personalization and billing flexibility, which leads to a win-win ecosystem. The underpinnings of all of this have been IP networks. So we’re excited about what some people call 4G networks such as LTE – and some would lump WiMAX in there as 4G too. WiMAX is already here and is being deployed, while LTE is being talked about, with trials scheduled for 2009, and rollouts occurring in the 2010 to 2012 timeframe. Regardless of when the LTE rollout occurs, it’s fair to say that end-to-end high performance IP architectures and solutions will drive the next wave of the Mobile Internet. That in itself is good for the ecosystem and Cisco, because we can build on our strengths and extend our portfolio of network elements to adapt to the needs of increasingly high-performance networks.”

 

“Also, if you believe that the Mobile Internet data explosion will continue, there is an obvious ‘choke point’ at the points of radio aggregation,” says Nagesh. “All of these handsets and other devices are now pumping so much data into the network that the scalability and performance aspects of the networks are extremely important. We do believe that in the Mobile Internet world, some providers and operators will successfully exploit these applications and their associated technologies, while others may ‘bite the dust’. So in some sense it’s a market inflexion point for how operators and service providers capitalize on these phenomena, adapt and build the appropriate infrastructure, intelligence and personalization. At times I’ve said that things will become so ‘micro-segmented’ that the network will capable of catering to individual users and SMBs rather than types of segments or demographic groups, so there is the possibility of exploiting the ‘long tail’ of the market using a nicely architected, end-to-end IP environment if there is sufficient network intelligence to support such personalization.”

 

“In terms of radio aggregation, in the 2G world, most of the mobile applications were voice. Your handset talked to a base station, which in turn would use a copper T1 or E1 which would take the traffic to the IP edge, and thence to an IP core,” says Nagesh. “Often many of these networks had MPLS-based IP networks which kept their quality in good shape. So in the voice world overall, the upstream connections from the cell sites were largely T1s and E1s. In the new Mobile Internet world, T1s are just grossly inadequate. You see an obvious choke point there. We feel that Carrier Ethernet-type services will step in to handle this, and high-performance cell site routers will become the norm, so that the operators and providers can innovate in terms of their business models. They could resort to models such as ‘all you can eat’ or ‘application-based’ or ‘personalized’ or whatever. In any case, the architecture needs to address the choke point to facilitate innovation in terms of business, billing, customization and personalization. You’ll see the IP network edge becoming smart, and the IP core will achieve greater performance and scalability. Cisco feels this is good, because it’s driving by consumer behavior and it promotes more interoperability and innovation, which is what IP has engendered throughout its existence.”

 

“In this world of the Mobile Internet, we feel the notion of mobility needs to be beyond just radio,” says Nagesh. “Sometimes Mobile Internet may seem to be an oxymoron. It’s Internet where you as a consumer can get what you want with any device anywhere at any time. That’s where it’s heading, but let’s assume for a moment that the next-wave comprises higher performance, rich media applications, and collaboration apps, all leading to higher performance radio networks. We feel that ‘mobility’ will transcend just being associated with a handset. It will involve, for example, scenarios such as you sitting at home watching CNN and you realize you need to go to work. You can just switch your communications nexus over to your iPhone, and the system knows where you are and the same CNN channel is being seen and heard by you as you walk to the bus or train station. That kind of situation is also included in our definition of mobility, and we’ve spent several months explaining this redefined notion of mobility, which is ‘any time, any place, any device’. These ideas can be extended to video, unified communications and so forth. You’ll see Cisco launching applications to reflect ongoing consumer and business behavior, which we call the Connected Life.”

 

“So the phenomenon is happening, but how will providers monetize the customer’s experience in a meaningful way while creating an architecture that gives them openness, flexibility and velocity of innovation on the service side, business side or application side?” asks Nagesh. “That is the framework in which I want to discuss what’s happening with wireless networks.”

 

“Radio backhaul has not taken off in the U.S.,” says Nagesh. “If you assume that the operators already have TDM or ATM circuits and a bunch of T1s or E1s, one possibility is to add more T1s or E1s, so the radio access can terminate gracefully at the cell site and then the IP network can take it to the rest of the network and then to the Internet, where the services, servers and peer-to-peer aspects can be found. In some instances 802.16d type products, microwave or other alternatives may useful, but the transport itself ultimately becomes IP, and Cisco has spent considerable time, money and energy to create many products that reflect these multiple forms of transport over IP, particularly the MToP part of it.” [The Mobile Transport over Pseudowires, MToP, is the Cisco solution for converged packet-based Radio Access Network or RAN aggregation between edge aggregation sites and the network core. MToP uses standards-based Multiprotocol Label Switching, or MPLS, technology to extend the packet-based core already deployed at many carriers to the edge of the network. This solution is flexible, cost-effective and allows operators to evolve to a packet-based network while still maintaining the easy operational familiarity and resiliency of the existing TDM-based network.]

 

“Many trends I’ve mentioned are indeed leading to changes in both RAN aggregation and RAN optimization in instances where the operator just wants to optimize what they have,” says Nagesh. “It also leads to situations where an operator wants to leverage Carrier Ethernet, or resort to architectures where one uses a 2G or 3G network to send only the voice traffic over it, and shift all of the data traffic onto the Ethernet side, whereupon one can now implement signaling, clocking, and so forth using pseudowires. I don’t want to get into the specifics, but there is a way to maintain the graceful migration wherein your voice traffic continues to do what it was doing, and you just shift the data to the Ethernet side. That in some instances may yield a better short-run cost of ownership with a clear path to the future, where it becomes easier to manage SLAs, interoperability and the type of metering they want to use. So the hybrid approach, with the combination of pseudowire and TDM, is being deployed by some of our customers. There are also customers who employ native IP backhaul for both voice and data, which is probably the most graceful approach that will be able to handle what is to come. We see a lot of traction in this area. Most recently we introduced our MWR2941 that can collect cell site traffic and send it over MPLS and Ethernet. Our Carrier Ethernet solutions also support IP over DWDM, and those type of interfaces, which really positions high per performance Ethernet services that can then integrate seamlessly into an IP edge and/or an IP-MPLS core to create an end-to-end mobile transport-over-packet scenario. We can do that easily.”

 

“On the radio side, the success of WiMAX, LTE, or whatever acronym you prefer depends on what spectrum you own, how you want to capitalize the spectrum you have, what your starting point is, and where you want to go,” says Nagesh. “Those are in simple terms the parameters that will decide your path. For example, if you are already on a CDMA network, you might go with EVDO. If you’re working with a GSM network, you may look at HSPA and HSPDA-type solutions. There’s still room there to make money, and despite all of the talk about LTE, most of the wireless networks are indeed 3G networks based on CDMA or GSM. They will continue to evolve towards the data domain – the ARPU is sort of saturating and the data traffic is rapidly increasing. So we feel that it is a strategic imperative for the provider to decide what their evolution will be in terms of a mature model of mobility, and my definition of mobility goes beyond just handsets. They also need to figure out how to bundle apps, what to do with ‘over-the-top’ applications, and how they will build the necessary intelligence in the network so that they can shift the slope on the ARPU curve. Interestingly, in some emerging markets, providers have lower ARPUs, but they have some interesting deploy-build-transfer models where the risk/reward has been shifted so that these providers have been profitable, even at lower ARPUs. But if you take the general industry at large, it is fair to say that whereas the bulk of the traffic is data, the bulk of the revenue isn’t derived from data, because that’s the ‘journey’ the providers are still taking. Voice is still a major phenomenon.”

 

“If you look at your own assets as a provider, you can see a natural path to migrate from a 3G network, say, to LTE – whenever LTE materializes, of course – but you can continue to exploit your 3G network to drive the data services and even gain higher performance with LTE’s promise of 100 megabit per second bandwidths,” says Nagresh. “That’s a natural evolution, and no wonder they call it LTE – Long-Term Evolution. The challenge for the industry is to answer questions from providers such as, ‘I spent so much money on my 3G network, I haven’t made enough return on my investment. And yet I’m now talking about moving to LTE.’ Another challenge involves the pilots and trials leading to scalable deployments. I’ve seen and read predictions about deployment timeframes ranging from 2011 to 2012. So we’re talking about a slightly longer term for LTE, which means that 3G networks will be the predominant data network on the mobile side.”

 

“As I said, things depend on what kind of spectra you as a provider own,” says Nagesh. “If you have the spectrum allocated for WiMAX, whether it’s 2.3, 2.5 or 3.5 gigahertz, why shouldn’t you capitalize on it too? End-to-end WiMAX solutions are offered by many companies, including Cisco. WiMAX and LTE can be lumped under the 4G rubric, under the premise of certain attributes and developments such as end-to-end IP architectures and solutions, rapid innovation for new applications that go beyond the normal SMSs and VPNs and into the realm of higher-performance and higher-intelligence apps such as unified communications, mobile video, and so forth. There’s enough bandwidth to make all of these things happen. On the business side, this will all drive new levels of productivity. We at Cisco are looking at some of these advanced apps for our WebEx Connect portfolio of products which can support both wired and wireless environments. You will see new apps for 4G, which will be good for both Cisco and the industry in general. Cisco has historically been radio-agnostic and will continue to be, even though we have WiMAX as an instance of a true end-to-end solution. But looking at the rest of our portfolio regarding backhaul, the IP edge, all the services and everything up the chain to the CRS-1 – those components are common to both LTE and WiMAX and even other types of networks. There is no ‘wall’ between LTE and WiMAX. The appropriate gateways will evolve to handle the special requirements of the different radio technologies. We already have GGSN PDSN gateways [The GPRS Gateway Support Node acts as an access point between a GPRS network and the Internet or private networks. A Packet Data Serving Node is the gateway between a service provider’s IP network and a mobile phone] and we have our WiMAX ASN gateway, which can be a blade that can slide right into our 7600 router, and it’s one of the highest performance gateways you’ll find. That portfolio will continue to evolve to support other radio technologies including LTE.”

 

The Almost-Here-Gen Network

 

The next-generation network is still to some degree on the horizon, yet many components of it are now appearing piecemeal in the networks of operators, carrfiers, service providers, MSOs, etc. Over the next few years we will be inundated with more advanced services and devices capable of running them. Fortunately, it’s not so much a huge bet on future technology so much as it is an attempt to cater to what both the business and consumer worlds are clamoring for.

 

Companies mentioned in this article:

 

Cisco Systems

www.cisco.com

 

Comptel Corporation

www.comptel.com

 

Oracle

www.oracle.com

 

NextPoint Networks/GENBAND

www.nextpoint.com

 

Starent Networks

www.starentnetworks.com

 

Tekelec

www.tekelec.com