Cisco has its eyes firmly on the mobile provider space (although not, as we discuss later in the interview, on providing the radios). Its aim is to own the mobile core, from the moment traffic leaves the wireless domain and hits the wired, all the way to the (presumably Cisco) core IP networks.
“From the router at the radio, to the aggregation, to the core, to the data centre – we can offer the operators a comprehensive network,” Capener said.
The networking giant’s pitches will be capacity, manageability, and provider revenue – all of which it hopes will be facilitated by the ASR 5000 system brought into the Cisco family from the Starent acquisition.
Wireless Data Growth
Regarding capacity, there’s no doubt about the accelerating growth in mobile data. Leaving aside the rest of the world, Australians are adopting 3G data services at a remarkable rate. Telstra alone added just over two million 3G users in 2008-2009, more than one-quarter of which were 3G data cards.
That growth, as Capener explained, is reflected internationally, with Cisco’s Visual Networking Index predicting that by 2014, “we expect that two-thirds of the data traversing the mobile network to be video, and the number of connections by 2014 will exceed five billion.”
With video as the driver and IP as the core network, Cisco sees a huge opportunity to leverage its IP expertise into the network infrastructure.
“If we look at the mobile
“And the operators have to increase the revenue.”
Handling the increase in capacity isn’t just a matter of building bigger pipes, because of characteristics that are peculiar to mobile networks. Mobile broadband carries a heavier signaling load, so the operator doesn’t need to cope with just the number of calls on the network, but also the “call transaction rates” (in other words, the ability to accept a new data call, set up the new session, capture the billing information, and end the call; all of which reflects the voice heritage of mobile networks).
Intelligence, he said, is also becoming a “hot button” in the mobile environment, because he believes it’s seen as the secret both to lowering the cost of offering mobile broadband as well as creating new revenue opportunities.
“You need subscriber intelligence, and you need network intelligence.”
The “subscriber intelligence” Capener refers to is the ability to know what subscribers are doing so the provider can offer the best path to the content; while network intelligence gives the operator the ability to understand traffic conditions and the sources and volumes of traffic at any given time.
Operators can use the combination of subscriber and network information, he said, to address issues of traffic optimisation and revenue generation.
The Need for Intelligence
Since so much of the history of the Internet has been based on removing intelligence from the network (the “smart endpoints, dumb networks” approach to network architecture), Search Networking thought it reasonable to ask why mobile networks seem to be going against this trend and seeking more intelligence.
It comes down to the constraints of the network, Capener said – some of which are technical and others administrative.
“Access to spectrum and network bandwidth is a finite element, and with mobile operators, it’s something they’ve invested billions in.
“So they need to more efficiently deliver the traffic, so as to optimize that investment – especially when we look at the amount of video delivered over the network.”
And there’s the simple and obvious issue of mobility: “People are moving about, which leads to varying network conditions. Sometimes, one particular cell will suffer from congestion – and you have to be able to deal with that.”
So from the network point of view, Cisco’s aim is to make delivery of traffic more cost effective.
The other side of the “intelligent network” coin is revenue generation – and here, the need is two-fold. Not only do operators already have to cope with traffic that’s growing faster than revenue, but increasing the intelligence in the network also counts as a cost.
Only, however, a relatively minor one, according to Capener: “The majority of the operator’s investment is in the radio.” Hence, he argues, “spending a little bit more … gives you service and optimization possibilities at a relatively low cost.”
About That Revenue Question
Cisco actually faces a formidable challenge in pitching itself into the mobile infrastructure space. In its home market of IP routing, Cisco is the exemplar of incumbency; but in mobile networks, it faces a host of incumbents (Nokia, Qualcomm, Motorola, NEC, Ericsson, Siemens … the list goes on).
So Cisco needs not only to reinforce its standing in IP routing (a simple enough proposition), but to convince them that it has the answer to their traffic-revenue equation.
One way is by offering solutions that can respond to the kind of device that’s consuming the video traffic Cisco’s Visual Networking Index tags as the category-killer – video.
“With the right intelligence, you can know what type of device is accessing the video, and therefore the quality of video you require,” Capener explained.
Since the capabilities of a smartphone are different from the capabilities of the laptop, the operator has the chance to manipulate the video traffic on its way to the user, and “optimise it for the device.”
“That means you are preserving some of the bandwidth, because you don’t have to send as large a file to the smartphone as if you were connecting to a laptop.”
And a more intelligent network, he said, gives operators the chance to offer tiered services to consumers, for example charging more for delivering higher-quality video.
In the other direction, operators might also find themselves able to cut deals with content providers as well. Capener’s example is Netflix: “to provide a higher-bandwidth or higher-quality service, you may go to the video producer with a revenue-sharing arrangement.”
(It’s an approach that reflects the different ways consumers react to questions of network neutrality, depending on the network they’re using. Propose differential service agreements on the fixed Internet, and you can expect an uproar; mobile broadband offerings are given much more latitude.)
“Another really simple example is content filtering – providing something like parental control, so that parents can limit what their kids are able to access on mobile device. And the operator can look at putting that capability in the network, where they see the traffic, and they might pay a per-session cost for that – they can get a very simple and easy business model.”