Carriers are between a rock and a hard place when it comes to soaring demand for mobile data, and they’re going to need a quiver full of arrows to address the issue.
It’s no secret that operators’ networks are or will be strained by the massive volumes of data coursing through them. What’s less clear is what they’re going to do about it.
As Deloitte put it in a recent report, carriers are in a difficult predicament: Demand is so high that any notable increase in capacity will be overwhelmed by data-hungry consumers. To illustrate the extent of the issue, consider this: AT&T’s data traffic has gone up more than 5,000 percent over the past three years. Those figures are echoed by dramatic findings by Cisco, which predicts worldwide data traffic to hit 3.6 Exabytes per month by 2014.
How to address this growth, particularly in the interim period before operators get their 4G networks established, is perhaps the most important issue facing operators today.
WHY NOT CHANGE PRICING?
The easiest – and least likely – way for operators to control traffic is to start charging people for what they use. The problem is well illustrated by usage patterns on AT&T’s network, where just 3 percent of users generate 40 percent of traffic. Thanks to flat-rate, allyou-can-eat billing, some people are consuming more than their fair share of data.
These users are making the best of buffetstyle smartphone dining at the expense of the network, says Phil Asmundson, vice chairman of telecom, media and technology for Deloitte.
“When one percent of users comprise 21 percent of all bit consumption, the all-you-caneat model doesn’t work – especially when it’s so hard to predict network usage,” he says.
The problem is best highlighted by a comparison of data usage by various subscribers done for Consumer Reports. The study found that the average iPhone user consumes 273 MB of data per month, almost twice that of the 150 MB consumed by other smartphone customers and several times the 54 MB used by BlackBerry subscribers. Furthermore, the study found that 12 percent of iPhone users use at least 500 MB per month and a third of them use more than 1 GB of data.
Asmundson says tiered pricing models could go a long way toward moderating data usage by heavy consumers and argues that the average consumer would be better off on a tiered plan.
However, carriers appear hesitant to move in that direction due to widespread consumer sentiment in favor of all-youcan-eat plans. Asmundson says subscribers are poorly educated about the amount of data they consume. As a result, they believe an unlimited plan is the best deal.
AT&T and Verizon Wireless seem to be taking this quirk of the consumer consciousness into consideration as they grapple with the issue of soaring data traffic. AT&T Mobility executive Ralph de la Vega has said the company has not yet decided whether to use tiered pricing. Verizon Wireless has sent mixed messages on the issue, with its Chief Technology Officer Dick Lynch saying at one point that the company was considering usage-based pricing and its CEO Ivan Seidenberg saying it would focus on mobile data bundles.
It’s not that carriers aren’t experimenting. AT&T took a baby step in the direction of tiered pricing with a $15 premium on the unlimited data plan for Apple’s iPad. Some companies have embedded pop-up windows into dongle software to tell users how much data they have consumed in an effort to better educate subscribers about data usage.
If implemented, tiered pricing could deter data hogs and make rates more fair for the average consumer. However, carriers appear to be hesitant to make changes to legacy service fees lest they anger customers. Also, it’s not in their best interests to potentially limit usage with data caps – it’s tantamount to limiting the growth of their business. As a result, operators will have to turn to alternate solutions to satiate what appears to be an unquenchable thirst for mobile data.
BANDWIDTH & THE BAG OF TRICKS
Unfortunately, there is no silver bullet to solve the capacity crunch, at least not until 4G is up and running en masse. Operators have to turn to a veritable bag of tricks to bulwark their networks against surging data traffic.
“[There’s] going to be a quiver full of arrows to address the issue because not one solution is going to solve it,” says Jennifer Pigg, vice president in Yankee Group’s Anywhere Network research group.
She emphasizes that service providers should be planning their migration away from legacy TDM and SONET SDH technology. “It’s an inefficient use of bandwidth and they can’t afford it, but they cling to it like a safety blanket because it gives them recoverability, reliability and integration into their OSS/BSS and provisioning systems,” Pigg says.
She also says using a TDM network to support data does not make sense in the long run given the need to run networks as efficiently as possible. “It’s obscenely wasteful in terms of bandwidth and capital expenditures,” she says. “They’ve got to move away from circuit-switched solutions.”
ABI Research analyst Aditya Kaul agrees, calling TDM “not the most efficient mechanism,” especially when used in backhaul.
“It’s important for operators to move towards IP backhaul,” he says. “You can get all the spectrum you want, all the additional base stations you want, but if your backhaul continues to be based on TDM, you won’t get very far because it [creates] a bottleneck.”
Operators are moving on the backhaul front. Verizon Wireless plans to leverage its extensive FiOS network to cover as many base stations within its ILEC footprint as possible with fiber, using Ethernet for transport. Verizon’s co-parent, Vodafone, plans to use microwave technology that will carry high-speed Ethernet traffic. Clearwire is using microwave backhaul over its all-Ethernet WiMAX network.
Carriers also are embracing so-called “smart offload” technology, which uses solutions like Wi-Fi and femtocells to relieve network congestion. AT&T has announced tentative plans to roll out more free Wi-Fi hot spots, a move that could help free up bandwidth in its cellular network. It is conducting trials of its MicroCell femtocell, which could help alleviate bandwidth congestion if widely deployed.
Kaul expects other operators to follow AT&T’s lead on Wi-Fi offload and says femtocells may become increasingly popular as a method of offloading traffic from an operator’s wireless networks.
Wi-Fi and femtocells are not the only techniques being used to address the capacity crunch. Making sure that traffic is being transmitted as efficiently as possible is also a crucial part of network management.
OPTIMIZING FOR BETTER MILEAGE
Several major vendors, including Motorola and Alcatel-Lucent, and smaller companies like Bytemobile are getting into the network optimization space as part of their efforts to respond to the capacity crunch.
Motorola’s Director of Market for Wireless Broadband, LTE and WiMAX Tom Gruba says that network optimization technologies can help operators squeeze extra juice out of their networks. By improving the energy efficiency of outgoing traffic and increasing the sensitivity of receivers, operators can improve the metaphoric “gas mileage” of their networks, he says.
“At a minimum, you want networks to operate as efficiently as possible,” Gruba says. “Operators want to make sure they say yes to every customer.”
Bytemobile’s Vice President of Product Management Joel Brand says network optimization technology can also improve the end-user experience. “It helps [operators] service more users on the given infrastructure; balance interference between users and applications; and fight congestion,” he says.
Technologies like deep packet inspection can help stop a small percentage of heavy bandwidth consumers from disrupting network traffic, although operators have expressed some concern over using the technique to monitor traffic. Compression technology also comes in handy, especially for data-hogging video traffic.
LOOKING AHEAD
At the end of the day, however, it all boils down to hardware. “There is no substitute for good engineering. You must be mindful of your network and you must take care of it,” says Iyad Tarazi, vice president of network development and engineering at Sprint.
Tarazi, who is responsible for the core engineering functions of Sprint’s network, says optimization technology has only an incremental impact on network performance. He also says methods like compression and deep packet inspection are “used very sparingly… We don’t subscribe to a model that limits customers’ access.”
Tarazi says his strategy for managing the traffic coursing over Sprint’s 3G and 4G network is to stay alert to rising demand. “You have to constantly get ahead,” he says. “You must forecast, predict and handle traffic before it hits you.”