4G Mobile’s Core Difference
VS: In broad terms, what the router vendors are doing is to take an existing router platform and then graft a packet core onto it. This creates some architectural issues in terms of scalability and performance. A router is a Layer 3 platform, while a packet core is
more of a Layer 4 through Layer 7 platform. When your initial product wasn’t designed to be a gateway product, it creates fundamental limitations when you try to add gateway functionality to it. However, when you start with a packet core platform and
try to add routing and switching — which is the approach we are taking — you get a completely different product with significantly better scale and performance characteristics. In some ways, both approaches combine routing and
packet core functions in one product, but the resulting architectures and designs are very different. We’re starting at the gateway layer and moving down the stack rather than starting at the routing layer and moving up the stack. The router vendors are building their packet
core on top of their existing operating system, which is built around routing and switching. In that model, when you add packet core you’re adding a server blade: an application layer blade in the router. You’re taking traffic in from a line card, hair-pinning it into a server blade and then bringing it out through another card. That is very inefficient because traffic has
dimension of performance. We’ve done things that we would consider more 4G-centric, rather than trying to reuse older platforms originally designed for 2G and 3G networks of the pre-smartphone era.
KF: What do you think the competitive landscape will be like? Many analysts have projected that a few radio access vendors will dominate the 4G radio market. Will the same be true — or even more so — for the 4G core market? VS: The closer you go to the edge, there is more opportunity for more players. As you get to the core, it condenses down to fewer players. I think most operators are drawing the line between
the radio and the core. Radios are very cost-sensitive components, and that’s where most of the CapEx lies. It’s good to drive the cost down by having more competition in that space. The packet core is a little different because
“ You want to make the user experience consistent.”
to traverse the switch multiple times. What we do is perform all the processing functions when traffic enters the line card. We have a very distributed architecture, in which we’ve
— Vikram Saksena, CTO, Tellabs
it’s the control point for the users’ experience. There are very few operators who will deploy multiple different packet cores. Each aligned with a different radio vendor, because when you move from base station to base station and end up on a different packet core, service layer consistency may be lost. In that situation, it’s not clear that you’ll
get the same user experience because of the different ways these packet cores deal with traffic and user sessions. Many operators are trying to create more of a single- or dual-vendor packet core, along with a multiple-vendor radio access
built the packet core into our platform operating system itself, as opposed to adding it as an application on an existing operating system. When the packets enter the line card, they get processed, not only at Layer 2 and Layer 3 but also at Layers 4 through 7. This reduces latency and increases the effective throughput of the platform. Compared with stand-alone packet core platforms, we
paid special attention to the multi-dimensional aspects of packet core performance. Our architecture supports independent scaling of control plane, data plane and application plane performance. The stand-alone packet core platforms from other vendors
were designed originally for 2G and early 3G networks. This was well before the advent of smartphones. Back then, data traffic demands were not as stringent, and only the data plane throughput was of primary concern. We recognized that with the plethora of smart devices coming on to the 3G and 4G networks, dealing with control plane and application plane performance was equally important. Therefore we dedicate processing resources to scale every
network. You want to make the user experience consistent, and you want to make your charging and billing functions consistent, so you won’t see as much diversity in the packet core as you will in the radio network.
KF: Are you saying operators are willing to pay a premium for a packet core that they wouldn’t necessarily pay for the radio network? VS: In our industry, generally speaking, pricing is value- driven. The packet core is the “brain” of the wireless network; it is where the service logic and policies are enforced, which enables the operators to differentiate their offerings. This is quite different from the radio layer, which
provides commoditized access. The pricing models are consistent with the value delivered by these segments in the operator’s network. n
2G: Second Generation 3G: Third Generation 4G: Fourth Generation
TELLABS INSIGHT Q3 7
CapEx: Capital Expenses RNC: Radio Network Controller
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36