Mobile Data Networks Face New Strain

Jonathan Christensen eBay Skype division general manager, says “the phone is dead," arguing that VoIP over mobiles will accelerate the trend. Agree or disagree, mobile network operators,will face issues other than loss of lucrative long distance calling revenues and bandwidth consumption, as the "VoIP over mobile" trend gathers speed.

As it turns out, says Mike Schabel, Alcatel-Lucent general manager, bandwidth consumption isn't the only problem mobile networks face, and in some ways may not be the key problem posed by IP applications.

Consider Session Initiation Protocol. As SIP applications start to represent more of any mobile user’s total usage, the problems become evident. Ignoring for a moment the revenue implications, SIP-based applications present previously-unacknowledged issues. The reason is that although SIP-based voice and communications are not a terribly big consumer of bandwidth, they are a huge consumer of radio network signaling resources. And it is radio network contention that is the gating problem, in some ways, not bandwidth consumption.

Where a typical user might place most of the bandwidth load on the radio network by using Web browsing, P2P and WAP applications (and where SIP bandwidth is negligible), the signaling load is highly disproportionate.

Where HTTP use might represent 44 percent of total bandwidth use, and consume 1.3 hours of airtime, while imposing 240 signaling events, a SIP application might chew up 3.9 hours of airtime and 2,240 signaling events despite using just 0.02 percent of total bandwidth.

Likewise, a VPN connection might represent just 2.4 percent of bandwidth consumed, but represent 20.75 hours of airtime, while imposing 5,970 signaling events.

So the problem for a mobile network provider is not simply a cannibalization of current revenue, but dramatically-more-intense pressure on the radio access network. And the issue isn't bandwidth: it is signaling overhead that chews up radio network element capacity, even when bandwidth is hardly used.

Necessity Drives Strategy at Qwest, Other Firms

Service provider strategy sometimes is dictated by necessity, and to the extent that service providers large and small now face different "necessities," there is an increasing divergence in strategy. Over time, in other words, service providers will "look" different from each other where in the past they all had resembled each other to a striking degree.

Consider Qwest, one of the three former "Baby Bells." Qwest always had a customer geography significantly more rural than the other original seven Bell Operating Companies. But when SBC gobbled up Telesis, Ameritech, AT&T and BellSouth to form at&t; while Verizon was formed from the former Bell Atlantic, NYNEX, GTE and MCI, the differences grew.

Both at&t and Verizon have the leading mobile assets and much-larger scale than any other contestants in the marketplace. Qwest is far smaller, does not own a national wireless network and faces much-larger challenges in the fiber-to-customer area because of lower density serving areas.

So where a triple play offered over owned facilities strategy makes sense for Verizon and at&t, for Qwest it does not. Qwest simply doesn't have the customer volume, density or access to capital that strategy would require. Unlike the other former Baby Bells, Qwest's fiber-to-customer strategy is not anchored by video services, but strictly by broadband data services.

Lacking a mobile network, Qwest originally tried offering services under its own brand, as a mobile virtual network operator. But it now has decided that approach has drawbacks, including some handset limitations, financial returns limited by low volume and, arguably, the lack of a popular "brand name" in wireless.

Qwest also has to maintain a balance between capital investment and shareholder return issues, such as reducing debt load, buying back shares and supporting the payment of dividends.

So Qwest's strategy will embrace partnerships in areas such as video and wireless, in ways that Verizon and at&t will not. In the process of revising its mobile strategy, Qwest also says it will rely on DirecTV for the video services component of its offerings. And where video services will be a key part of the payback for FiOS and LightSpeed, Qwest expects to get its payback strictly from new broadband services.

That's going to necessitate high penetration and new services as well. By 2011 Qwest plans to increase its broadband penetration from 23 percent to 40 percent.

Qwest will "look different" in its strategy because it has to. It doesn't have the scale or resources to become a smaller version of at&t or Verizon. Consider that at&t books about $39 billion annually while Verizon books $24 billion annually.

Qwest books about $13.8 billion a year in annual revenue. Neither does Qwest closely resemble the middle tier of independent telcos, either. Embarq, for example, books about $6.4 billion in annual revenue. Windstream books about $3.3 billion annually. But most independent telcos are far smaller than that, booking millions to hundreds of millions worth of revenue each year.

What makes Qwest different from the mid-tier of telcos such as Windstream and Embarq is that Qwest operates global backbone networks that can feed a more-robust enterprise business. The other providers might more logically be called regional "local" providers. Some competitive local exchange carriers also have a "national local" character, the difference being that such firms generally only serve the business customer segments.

In many ways, the interesting strategic paths will be among the smaller telcos rather than the tier one providers. Very-small independents typically are very interested in offering IPTV services. The middle tier of companies generally are not. The middle tier of companies may have brighter prospects in business customer segments. Very-small providers typically will not. Very-small telcos may not find out-of-region operations too compelling. The middle tier, at some point, virtually has to look at footprint expansion.

Since strategy is the result of multiple background, financial and management factors, we can expect that some of the more-differentiated approaches will be taken by those providers who are particularly challenged in some way. Inability to create the triple play or quadruple play strategy; geographic or demographic limitations or sheer borrowing power will force some managements to strike out on atypical paths from that generally seen as the tier one global provider approach.

at&t to Raise Text Messaging Rates

It appears at&t Wireless will be raising the prices it charges for casual sending and receiving of tex messages, though "bucket" prices remain unchanged. Effective at the end of March, the charge for text messages not part of a messaging bucket will be 20 cents.

For most of us, that means the buckets make even more sense. For light users, the move just shows how data services are becoming the revenue model for mobile services, with voice gradually declining in importance.

WiMAX Segmentation

One of the unresolved questions about WiMAX networks in the U.S. market is whether a sizable new business can be created around devices other than cell phones and smart phones.

And there is at least some reason to believe an opportunity exists, though pricing might be an issue.

Analysts at Compete Inc. recently asked consumers shopping online for consumer electronics devices about their interest in connecting devices to the Internet.

More than 50 percent of laptop and GPS shoppers were very interested in devices that enable enhanced connectivity using an open access network.

A follow-up question revealed that consumers are also willing to pay for this connectivity, with about 25 percent willing to pay over $50 at the time of purchase to include this feature, Compete suggests.

Recurring costs are the bigger issue, though. It isn't clear how many users will be happy to pay recurring connection fees if the option to use their in-home Wi-Fi networks is available for no incremental cost. Up to this point, no matter what they might say as part of a poll, few camera users have proven willing to spend money for network services.

Sprint Mogul to Use Rev A Broadband

Sprint is releasing a software update for the Mogul phone, made by HTC Corp. of Taiwan, that will enable the phone to connect at Rev. A speeds.

Downloads speeds should be 600 kilobits per second to 1,400 kbps, up from a range of 400 kbps to 700 kbps with Rev. 0.

It will be capable of uploads of 350 to 500 kbps, up from 50 kbps to 70 kbps.

Traffic Shaping Enhances User Welfare

To protect all users of shared access resources from service degradation, it makes sense to charge a congestion premium or use traffic management techniques, say researchers at the Phoenix Center. When congestion-causing applications degrade the experience of other users, the most efficient traffic management actions would be targeted at applications that cause congestion externalities and not upon all applications generally, say George S. Ford, PhD, Thomas M. Koutsky, Esq.and Lawrence J. Spiwak, Esq., Phoenix Center analysts.

Ironically, service providers tend to do too little to reduce the harmful effects of negative externalities caused by network congestion, they say. So those who argue that the Federal Communications Commission needs to impose prohibitions against network management practices because broadband providers will always be “too aggressive” in clamping down on uses of their network have it precisely backward, the researchers argue.

"It is socially desirable to charge a congestion premium when congestion-causing applications are used on a broadband network," Ford, Koutsky and Spiwak say. That is especially true when the congestion charge targets a particular congestion-causing
application, not blanket "price-per-bit" rules, they argue.

Indeed, if such charges are not targeted, then the price premiums may not achieve their desired purpose, Phoenix Center argues.

The objective of such charges is to attenuate congestion by requiring users of bandwidth-greedy applications to consider more fully the
congestion costs imposed on others.

"The fact that a broadband service provider operator may engage in application-specific traffic management techniques should not necessarily be viewed by a policymaker as evidence of illicit anticompetitive intent, the researchers say.

In fact, congestion is more likely to occur in shared media networks, such as wireless broadband networks, where all users share the common pool of spectrum capacity.

The complexity of this issue indicates that specific, prescriptive rules that ban entire categories of traffic management techniques across all network architectures and topologies can result in sub-optimal outcomes, they say.

"Our focus is upon the presence of congestion externalities: that is, the use of applications by some users that reduce the value of broadband service to other users on the broadband network, without compensation, by causing delays or other service quality problems," the researchers say.

"In the presence of a congestion externality, network management—including, but not limited to, the differential treatment of particular applications—is welfare enhancing," Ford, Koutsky and Spiwak say.

Apple Won't Block VoIP Apps Using Wi-Fi

CEO Steve Jobs says Apple will not block third-party development of VOIP applications for the iPhone and iPod Touch, as long as the applications run over Wi-Fi. That's an obviously politically astute move given Apple's relationship with mobile providers who aren't keen on the idea.

Nevertheless, the idea might allow everybody to get a bit more experience with "dual mode" service, using a widely-deployed handset, without a lot of expensive bets on particular devices, interfaces or marketing.