New Verizon Policy on Heavy Users, Congested Towers

Verizon Wireless has instituted a new network management policy that some will call “throttling,” while others might say simply represents a more-nuanced way of managing network congestion.

The new plan affects what Verizon says is about five percent of Verizon’s user base, specifically those users of 3G services that use 2 Gbytes of more of data each month, from congested cell sites. The rules do not apply to users of the new 4G network, though. The easiest solution is simply to use 4G. It’s a better experience anyway. New Verizon usage scheme

One suspects that users are capable of making rational choices about their services, and also will rapidly adopt the “default to Wi-Fi strategy.” Most people already seem capable of quickly grasping the advantages.

Some 64 percent of smart phone consumers surveyed by Devicescape use Wi-Fi hotspots at least once a day. Most smart phone owners who use Wi-Fi also use it on the road.  The study showed 90 percent of those users report accessing Wi-Fi both at home and on the road. Smart phone users use Wi-Fi often

Of those who use Wi-Fi outside their home or office, most (24 percent) connect at a cafe or coffee shop, 17.3 percent at a hotel, and 15 percent at a school campus. See Facing data caps, consumers keep turning to Wi-Fi.

Historically, mobiles haven’t been used excessively for data connections. Average mobile data consumption increased from about 90 MBytes per month during the first quarter of 2009 to 298 MBytes per month during the first quarter of 2010, according to Nielsen.

This represents a year-over-year increase of approximately 230 percent, though.  While this increase is substantial, in the first quarter of 2009 more than a third of smart phone subscribers used less than 1 MByte of data per month and usage has dropped to a quarter in the first quarter of 2010.

About 20 million current smart phone users are hardly using any data.

But there is a reason frameworks for managing bandwidth use are important. As mobile data consumption continues to grow, the usage pattern is starting to resemble fixed-line patterns, and that is a problem for all mobile service providers, as there is not now, and never will be any way for mobile providers to match the bandwidth, or cost of bandwidth, that a fixed network provider can offer.

There is a telling statistic in Cisco's Visual Networking Index, namely that as mobile broadband users have rapidly grown, their usage pattern rapidly has assumed the familiar pattern seen in the fixed-line part of the business.

Consider heavy usage patterns. The top one percent of mobile data subscribers generate over 20 percent of mobile data traffic, down from 30 percent just a year ago. That 29-point swing in just 12 months suggests that as more "typical" users adopt mobile broadband, they bring behaviors much different from those of early mobile broadband adopters, namely less-intensive consumption.

Cisco also reports that mobile data traffic over the last year also now matches the 1:20 ratio that has been true of fixed networks for several years (one percent of users generate or consume 20 percent of total transferred bytes). Visual networking index

Similarly, the top 10 percent of mobile data subscribers now generate approximately 60 percent of mobile data traffic, down from 70 percent at the beginning of the year.

All of those instances of "reversion toward the mean" are driven by the broader adoption by "typical" users of smart phone service. That noted, average smart phone usage doubled in 2010. The average amount of traffic per smart phone in 2010 was 79 Mbytes per month, up from 35 Mbytes per month in 2009.

Bandwidth and Revenue: Something's Gotta Give

Fully 73 percent of industry executives surveyed by Telcordia expect to see either network-enforced hard limits, tiered services or a combination of both to manage data traffic.

The survey of industry professionals across more than 75 countries by Telcordia simply confirms that data traffic, and therefore network cost, is increasing faster than revenue to pay for supplying that bandwidth.

"CSPs need to insert themselves in the mobile broadband value chain and leverage billing and charging assets to manage network costs and to add innovative value-added offers and services to both subscribers and over-the-top content third parties," says Pat McCarthy, Telcordia VP.

"All-you-can-eat data plans are not a sustainable business model, and policy-based bandwidth management and real-time charging provide CSPs and their subscribers with the necessary flexibility to try new services while keeping costs in line," says McCarthy.

link

Mobile Broadband Will Need a New Business Model

One way or the other, something has got to change in the mobile business as voice ceases to be the industry revenue driver. Today mobile service providers get 86 percent of their revenue from low-bandwidth applications like voice and text. But that will keep changing in predictable ways.

Namely, most capacity requirements will be driven by low-margin data rather than high-margin voice and text.  Over the long term, it is irrational to better price services in relationship to cost without attributing more revenue directly to the data services that are driving capital investment.

That doesn't mean every single service or application necessarily has to be priced in relationship to cost. Loss leaders at supermarkets, promotional DVD prices at Target and other promotional pricing happens all the time, in every business. Some products have high margin, others low or even negative margins.

The point is that current retail pricing will get more irrational as data demand grows, and that something will have to be done about it.

Carriers are investing in new capacity, but that alone will not be enough to bring revenue and capacity into balance. By 2013, virtually all traffic load will be driven by broadband data of one sort or another, especially video. That means, over time, new ways of charging for network usage will have to be created.

Like it or not, network management is going to be necessary, plus traffic offload and policy management. The issue, in part, is that demand is unevenly distributed. Even at peak hours of congestion, only a minor percentage of cell sites actually account for most of the congestion. To speak of congestion management at the "whole network" level is not to capture the issue.

The key issue is peak-hour congestion at perhaps 10 percent to 15 percent of sites. Put another way, even at peak congestion, 85 to 90 percent of sites do not experience difficulty. That means it might be necessary to use different policies at a small number of physical sites, not the entire network, even at peak hours.

So even if traffic shaping, bit priority policies and other tools are not generally required at every site, for every application or user, there will be a need to do so at some sites, some of the time.