FCC Demands Real Metrics to Define Broadband

When thinking about a definition for broadband, entities should think beyond 2009 and draw a “growth curve for how this evolves over time,” said Robert Curtis, the FCC broadband team’s deployment director. Newman agreed: “We need to think about how to quantify the notion of an evolvable definition.” Newman said one mistake he made when he was FCC chief technologist was coming up with the “infamous” 200 kbps minimum definition of broadband. “But to defend ourselves, that was 1999 and we defined it as an evolvable definition,” and he “never dreamed it would [still] be 200 kbps a decade later.”

One way to craft a definition is to determine the “minimum set of applications” that government wants every U.S. household to have access to, said AT&T Labs researcher Paul Henry. For example, required apps may include basic Web browsing, education tools and convenient access to various government organizations like state departments of Motor Vehicles and the U.S. Internal Revenue Service, he said.

The FCC could base the definition “on some sort of parity among areas that actually are being served,” rather than “come up with actual numbers,” said IP Action Partners President Stuart Lipoff, a consultant to NTCA. Under such a plan, the goal would be to bring unserved people to a level consistent with the lower-tier offerings received by those considered served, he said.

The plan must include an evolutionary path, but setting a high goal is critical, said Geoff Burke, a senior director at fiber company Calix. “Either we're going to the moon or we're not,” he said. Unless one establishes a goal, “you tend to get into these incremental steps” that aren’t necessarily the most cost effective ways to get “where you want to go,” he said. Policymakers “should look at what’s the best and get us as close as we can get to it,” said David Burstein, editor of DSL Prime. “I look at not what’s the bare minimum, but what’s the practical thing … to deliver.”

Consumers care less about actual speed figures than what the connection can do for them, said Telecordia Chief Technology Officer Adam Drobot. The next generation of broadband should give people access to any material they want, when they want, he said. Further in the future, broadband should be “anticipatory,” giving relevant information to people before they ask for it so they know what to do ahead of time, he said.

An Internet connection should not dictate what applications a user runs, or what devices they connect, said Google Vice President Vint Cerf. “As we try to fashion policies with regard to broadband deployment, you should keep in mind that this unbound notion of access to high capacity is what enables all kinds of new opportunity,” he said. “I would be unhappy if our regulatory policies or implementation … constrains the flexibility with which we can actually use the broadband resources.”

Social value may be reason enough to fund broadband deployment in rural areas, Curtis said. “If you believe that that total value ecosystem is enough for somebody -- let’s say the government -- to subsidize because you're creating societal value, then theoretically there’s enough value in the system” to justify deployment in rural areas, he said. The value extracted from broadband is extracted by more parties than just the infrastructure provider, Cerf said. “The ability for the maximum set of parties to extract value from that infrastructure is what we should be going for.”

The U.S. should think of broadband as a “fundamental utility infrastructure,” said Columbia University Prof. Henning Schulzrinne. Policymakers should strive for a network that’s “invisible” to users. People think of other central infrastructures “as working best when you don’t have to think about it,” he said. “We have to think far too much about the network today as a normal consumer.” At the same time, more broadband provider transparency is needed so consumers can assess the advantages and disadvantages of competing networks, Schulzrinne said. It’s tough for consumers to get information on costs and performance, he said.

To encourage broadband investment, regulators should seek to erase uncertainty in the market, particularly around network neutrality, said Amazon Vice President Paul Misener. “We were concerned five to six years ago with the unintended consequences of regulation,” he said. “Well at this point, I think we're seeing almost the flip side -- the unintended consequences of no regulation.” The lack of clarity may be stopping new services, he said.

Regulation mustn’t restrain industry, said Cisco Chief CMTS Architect John Chapman. For example, policy shouldn’t block Verizon from getting into the content business or discourage cable companies from upgrading their networks, he said. Regulators “should do no harm” and avoid stopping experimentation, said Drobot. At the same time, broadband infrastructure must be “paid for somehow,” and leaving too much ambiguity can have the effect of stopping investment, he said.

Regulation can be nuanced so it blocks unlawful practices without restricting reasonable ones, said Cerf. For example, some “differential services” used to manage traffic “may not be anticompetitive,” he said. “They might simply enable certain services that wouldn’t work” otherwise. This isn’t a problem for legislators to solve, he said. “This is the kind of problem that you almost have to deal with on a case-by-case basis,” recognizing there’s a spectrum of management activity with “value” on one side and “hazard” on the other, he said.

A major theme of an afternoon panel on wireless technology was the need for more licensed spectrum, a recurring complaint of wireless carriers.

“It’s going to be essential that we have more spectrum on the market, large blocks of spectrum,” said Tom Anderson, head of architecture for mobility at Alcatel-Lucent. “Although a lot of work is going on in my organization and organizations across the globe to increase the efficiency of spectrum fundamentally we're not going to be able to keep up with the growth rates without more spectrum.”

Scott Corson, vice president of engineering, at Qualcomm Flarion Technologies, said getting more licensed spectrum into carrier hands is critical. “From a mobile broadband perspective the consensus view, at least at Qualcomm, is that unlicensed doesn’t cut it,” Corson said. “We can’t figure out how to make it work.” Technology is rapidly reaching a point that little needs to be done to develop faster networks or better devices for mobile broadband. “The technology doesn’t matter so much,” Corson said. “It’s simply about the availability of licensed spectrum in order to deploy the technology. The problem is there’s not enough. We need more.”

Sten Andersson, head of wireless networks solutions at Ericsson, predicted that the speed of mobile networks using HSPA will increase from 42 Mbps today to 1,000 Mbps in 2014. Spectrum is the key, he said. “The more bandwidth you have available, the higher speeds you can accomplish and you get better efficiency,” Andersson said. “Spectrum needs to be made available, somewhat urgently. It takes many years to deploy new spectrum as you know.”

Barry West, president-international at Clearwire, said people invariably will demand mobile broadband. “I remember very well people saying ‘why would you need a mobile phone? A fixed phone is just fine,'” West said. “Here we are 25 years later and I just about guarantee that everybody in this room and most people on the planet who can afford it have a mobile phone and would not give it up. … The next phenomenon is going to be the mobile web. The way people get to the Internet is going to be by mobile devices. Why? Because basically that’s what human beings want.”

Kris Rinne, senior vice president of architecture and planning at AT&T, said the national broadband plan should allow carriers to manage capacity on their networks, ensure that spectrum is available to meet carrier needs, reduce the potential for interference and avoid requirements that would make networks less efficient. Rinne said AT&T has “invested $20 billion in network upgrades, spectrum and acquisitions. We plan to spend $17 billion to $18 billion, probably the highest of any company in the U.S., to extend and enhance the wireless and the wired networks.”

Newman said the testimony of industry executives does not answer one key question. “Nobody gave a concrete example of where we get that more spectrum,” the FCC official said. “You want more licensed spectrum. Tell us what should be doing? Who do we take spectrum from to get more licensed spectrum?”

During a second afternoon panel, on technology and rural deployment, Jim O'Connor, director of CPE engineering and planning at Open Range Communications, said rural carriers face many of the same problems as large carriers. “Site acquisition in the rural markets is not easier or faster than in the urban areas, as you might think,” he said. “The approval cycle of zoning and planning commissions can actually be quite long if they're meeting on, say, a monthly basis and you have to go through several design iterations or several site selections.” O'Connor said access to backhaul also can be a challenge in small markets, because of low volume at cell sites.

Vanu Bose, CEO of wireless infrastructure solutions provider Vanu, called technology crucial to improving rural deployment. He gave as an example using microwave backhaul to keep down the costs of connecting cell sites to the network. Network sharing also would help drive down costs, he said. Eighty percent of major carriers’ revenue comes from 30 percent of their cell sites in places like New York, Boston and Miami, Bose said. “The flip side of that is the bottom 50 percent of their cell sites, which are the least productive, the rural ones, generate only 10 percent of their revenue,” he said. “Asking [carriers] to extend that model deeper into rural to improve coverage just doesn’t work.”