FCC Advisory Committee Proposes ‘Logo’ Program to Compare ISP Speeds, Price

Not everyone on the advisory committee supported the logo proposal. Charles Kalmanek, AT&T vice president-research, cautioned that by only reporting on speed and price, a logo could make a less-capable ISP look better to a consumer, without regard for intangibles like customer service. But unlike something like the voluntary Energy Star program that measures the energy efficiency of consumer products and appliances, ISP performance might be too complex to be accurately captured in a simple logo, he said. Alcatel-Lucent CTO Marcus Weldon warned that the proposed logo could be too simplistic to do much good. It would be like only putting calories on a nutritional label, he said: “Until you've educated people about the numbers that matter, you can’t have a meaningful label."

"It seems like a good space for more voluntary activity,” said David Clark, senior research scientist at MIT and vice chair of the advisory committee, who cautioned against anything “too rigid.” Committee chair Jonathan Zittrain, a professor at Harvard Law School and Harvard’s Kennedy School, urged the working group to be specific about what exactly the logo would accomplish. It “could occasion a non-regulatory push to a certain ideal,” he said.

European regulatory bodies have already done a lot of research into this, concluding that this kind of data is helpful to consumers, said Alissa Cooper, chief computer scientist at the Center for Democracy & Technology. “Providing a label is a means to meet that need, even if you don’t think the label accurately reflects the complexity of this product,” she said. “It may be the case that no matter what you do, consumers treat this as a simple choice” based on little more than speed and price. A logo could help them make the choice, she said.

The problem might be complex, but that’s also true in other areas where the government has asked vendors to publish more data, said Brad Burnham, founding partner at Union Square Ventures. “Avoiding publishing the data because it’s complex doesn’t seem to be the right answer."

AT&T, FaceTime

AT&T’s restrictions on FaceTime raise “several interesting issues,” the group wrote (http://xrl.us/boa9d9), such as whether pre-loaded applications raise bigger network management concerns than applications that require a manual download. Rapid adoption of a new application might lead to “large and unpredictable changes in the traffic load on a cellular data network, which might necessitate “limited trial deployment of a new application to better understand its effects before wide-scale deployment,” the group said.

The group agreed that blocking applications runs the risk of discouraging innovation, but that carriers also need effective ways to manage the limited resources in cellular networks, it said. From the perspective of application developers, AT&T “did not choose the optimal approach” by blocking FaceTime over certain data plans, the group said. “By singling out one popular application, the door is opened for carriers to block lawful use of applications, require customers to upgrade to potentially costlier, limited plans, and justify their actions by claiming to be engaged in reasonable network management practices,” it said. “Agnostic” management approaches should be exhausted before blocking specific apps, it said.

In contrast, from a carrier perspective there are “good reasons to be concerned” about network overload because of FaceTime and similar apps. In that respect AT&T’s approach of limited enabling of the application was “a reasonable way of managing the risk of network congestion.” From a network equipment vendor perspective, apps should use adaptive session control techniques to adapt to changing network conditions, the report said. “In this context, it is reasonable to conclude that AT&T, with the largest number of iPhone users and largest fraction of their subscribers as iPhone users would have particular concerns about the load that the Facetime application would put on their network, with the potential to significantly degrade the available bandwidth for all other applications."

Burnham cautioned that discriminating against specific applications would be “very bad for innovation” because then investors won’t invest, unsure about whether the application will ever get to market. There’s a “difficult tradeoff” between the consumer’s interest in innovation and having an unlimited plan, and the app developer’s interest in a nondiscriminatory environment where its application will get to market without being throttled or priced higher by a carrier, he said. The mobile working group wants to remove the “ambiguity” from the app developer as to whether there’s a risk in developing a product, said Clark.

FaceTime presented an interesting case because, in contrast to many apps, it has a heavy uplink usage and a “not terribly adaptive codec,” so it can create high quality calls, Cooper said. That catalyzed the carrier’s reaction, and how they handled it, she said. An application like YouTube, even if bundled with an operating system, wouldn’t pose the same network concerns because the majority of YouTube’s traffic is in the downlink, not the uplink, she said. Many factors go toward creating a “potential pain point” for the carrier; the question is whether the committee can develop an “agnostic solution that can endure” across different situations, she said. “We're looking for win-wins,” said Jennifer Rexford, professor of computer science at Princeton University. “If an application is more judicious about how it’s using the network, the application will be more widely used.” AT&T announced Wednesday it would allow FaceTime regardless of data plan (CD Jan 17 p14).

The group also pondered the implications of the net neutrality rules on application developers. One item that arose during discussion was what role and expectation there should be for applications designers, Clark said. On the Internet they can do pretty much whatever they want, but in the wireless space “it’s clear that there isn’t infinite capacity, or enough capacity that you could use it casually,” he said. The mobile working group is considering advising developers about how to design their apps so that they are “somewhat disciplined,” he said.

The specialized services working group spent a considerable amount of time struggling with what specialized services are. The Open Internet Order gives examples of user-facing services such as facilities-based VoIP and IP video. “'Specialized services’ is a term that is meaningful only within the context of the Open Internet Report and Order,” Clark said. The working group cautioned that regulation should not create a perverse incentive for operators to move away from a converged IP infrastructure. “A service should not be able to escape regulatory burden, or acquire a burden, by moving to IP,” he said.

Specialized services are things you otherwise cannot do under the open Internet rules, “but that should be done,” said Barbara van Schewick, Stanford Law School professor, who is not on the advisory committee, but helped develop the rules. Kalmanek said that under the current rules, “there’s an ambiguity in my mind as to what we are allowed to do with respect to quality of service in this space.” Kalmanek isn’t sure whether, for example, AT&T would be legally allowed to offer a quality of service option if requested by Vonage.