Policy and regulation

The National Association of Telecommunications Officers and Advisors (NATOA) has called for a range of policy changes and investments that includes a guaranteed right for local governments to invest in broadband and fiber as the preferred mode of access.

Comcast has been reprimanded by the FCC for blocking certain Internet services without telling customers and without regard for the level of use. Service providers can be as arbitrary as they like in managing their Internet access services, but they have to tell customer what they are doing. A service-based network would not need to penalize customers for doing something they enjoy, but they would probably pay more. Today's broadband model is upside down for service providers because they make the most money when a customer never turns their computer on; they make the least if the customer is on the Internet all day long. That is the exact opposite of every other business in the world, and it's why broadband is such a mess in the U.S. Fortunately, communities like Danville, Virginia are changing that model.

The FCC has finally released new definitions of broadband.

• First Generation: 200 Kbps up to 768 Kbps

• Basic Broadband: 768 Kbps to 1.5 megabits per second
• 1.5 Mbps to to less than 3 Mbps
• 3 Mbps to less than 6 Mbps
• 6 Mbps to less than 10 Mbps
• 10 Mbps to less than 25 Mbps
• 25 Mbps to less than 100.0 Mbps
• 100 Mbps and beyond

This is a major improvement over the old definition of "200 kilobits" as broadband. By this old definition, the country has very high levels of broadband penetration, but made the U.S. the laughingstock of the rest of the world. In much of Europe, residential broadband tends to be north of 40+ MEGABITS, or about 200 times more capacity than the FCC definition.

The graded scale is useful because it can provided benchmarks to measure progress in a community or region. If the FCC has provided targets, that would have been even better. For example, a ten year target could be to have 90% of businesses and homes in the "100 Mbps and beyond" category, and indeed, U.S. community broadband projects like the one in Danville, Virginia are deploying "100 Mbps and beyond" today.

Comcast has announced that it will start slowing down the traffic of its broadband users if they are using too much; "too much" generally means running P2P (peer to peer) filesharing applications like BitTorrent, which can run for hours or days while sending or receiving large files (like movies or music).

The strategy is reasonable, given that cable companies price their Internet services in part based on average use. Customers that far exceed those average use parameters slow things down for everyone else on that cable modem network segment, which often includes 100-200 neighbors. Cable modem bandwidth, like most wireless services, is shared among all connected users at an access point (wireless) or a cable network mode. Fortunately, not all users are doing something at the same time, but background applications like BitTorrent do run continuously for long periods of time.

FCC Commissioner Deborah Tate spoke on the last day of the Broadband Properties conference. She had some interesting statistics that should give pause to anyone who thinks that DSL and cable modem broadband services are "good enough." Commissioner Tate noted that:

• Demand for bandwidth has been doubling every two years for the last ten years.
• By 2015 (just seven years from now), the FCC thinks bandwidth requirements will be fifty times (50x) what they are today.
• In Japan, where they have had 100 megabit connections to homes and businesses available for several years, they are already observing congestion--meaning 100 megabit pipes are already filling up.
• Americans are watching more than 10 billion videos per month over the Internet.

Tate went on to enumerate that choice was important to buyers of telecom services, and she listed that choice should be available for services, for providers, and for equipment.

The undersea fiber cables that were cut a couple of months ago were the subject of numerous conspiracy theories, but satellite photos have revealed the culprits--cargo ships that were anchored in the wrong place. Sometimes Occam's Razor (the simplest explanation is the likeliest one) is exactly right.

The object lesson for communities is to plan for cable outages by making sure local networks have redundant cable paths. Sometimes this is quite expensive to do when just getting started with community telecom investments, so an alternative to a second fiber cable is a high capacity wireless link that can handle local traffic (perhaps with somewhat less throughput) while repairs are made.

Eliot Spitzer, the governor of New York, has called for universal access to broadband in the state. The text of his speech is here (note that you have to scroll down past the agriculture remarks to get to the broadband stuff).

Unfortunately, Spitzer seems comfortable relegating rural areas to second class status. He calls for a minimum of 100 megabit connectivity in urban areas, but says that just one-fifth of that (20 megabits) is fine for rural areas. Cable and DSL are not going to provide universal access in rural parts of New York, so Spitzer has apparently decided that rural areas will have to make do with wireless while the cities get fiber. Rural citizens and legislators in the state should be outraged that the governor is willing to choke their economic future so easily.

This news report suggests that some cable companies are actively blocking certain kinds of traffic on their networks. The target of such blocking is peer to peer file sharing, in which the subscribers are often sharing very large files like movies and TV shows. From a network operator perspective, what you see is a very small number of your customers using a disproportionately large chunk of your network bandwidth, which can degrade service for other customers and increase costs.

Of course, the whole issue of net neutrality comes into play here, where the argument is that Internet access providers should not block traffic or offer preferential treatment to certain kinds of data over other kinds. But the access providers are on the horns of a dilemma, because the peer to peer sharing does have a measurable impact on their network.

But almost all of the arguments for and against traffic blocking, the rights of access providers to manage their own traffic, and free speech issues are red herrings. The real problem is the outdated and obsolete business model that nearly every access provider is still using.

It made sense, more than a decade ago, to sell bandwidth by the bucket to users because all folks were doing back then was email and some light Web browsing. And the Web in the mid-nineties was most text and a few pictures. Today, the media-rich Web is filled with sound and movie files of all kinds, and everyone uses them--look at the popularity of YouTube.

The current business model does not provide any cost information to consumers; that is, the "cost" of downloading a few emails is exactly the same as the "cost" of downloading an entire movie in DVD format. This is a classic and well-understood economics problem. Current pricing, to users, makes it appear as if there is unlimited supply (of bandwidth), which in turn creates unlimited demand.

The solution is simple. Rather than trying to discourage customers from doing things they want to do, you change the pricing model so that it conveys more information about the real cost of downloading things like movies.

In other words, you move from a business model of networks based on bandwidth, which we know is not working, to a business model based on unlimited bandwidth but where pricing is based on the service you want to use. So users that like to do peer to peer file sharing subscribe not to a bucket of bandwidth but instead subscribe to a service designed specifically to support high performance file sharing, and the price of that service will be based on the real cost of providing the service.

It is simple economics. Communities like Danville, Virginia are already rolling out networks that do this. These multi-service, open networks support real competition and will start driving innovation on the network because the business model support innovation, rather than stifling it.

Fights over WiMax spectrum are slowing deployment of WiMax. The FCC, which manages the WiMax spectrum, has been renewing the existing spectrum, called EBS (Educational Broadband Services). The problem is that the EBS spectrum licenses, in many cases, belong to local educational institutions. Sprint wants to build a national WiMax network and thinks that the FCC should require the schools not using the spectrum to give it up.

To make things more confusing, Clearwire, another WiMax provider, has taken the route of simply negotiating licenses directly with the schools, who make some money from something many of them were not using.

The end result will be extensive overbuilding of WiMax networks, which raises costs and makes it more difficult for users to roam from network to network. Wireless broadband operators have never been able to work out roaming agreements the way the cellular industry did. Cellphones did not become popular until roaming agreements were in place, meaning your phone would work pretty much everywhere. Today, in most airports, as one example, there are often two to five WiFi providers, and paying for service on one operator's network does not let you roam on any other operator's network.

The bigger problem here is overbuilding. With several different companies all trying to build wireless broadband networks in a community, costs go up for all users because of duplication of infrastructure. The solution is for the community to build a multi-service network that allows multiple providers to use a single network. Users gain the benefits of true competition, and prices are lower because there is no duplication of infrastructure. The FCC could play a valuable role here by encouraging the development of multi-service networks, but instead, continues to try to put band-aids on outmoded policies.

A 1998 ban on taxing services provided over the Internet is due to expire next month. Congress has three options: make the tax ban permanent, extend the ban for several more years, or start raking in a whole new source of cash.

If Congress decides to tax Internet access, everyone's access provider bills (dial up, DSL, cable modem, wireless, Blacksberry, etc.) could jump as much as fifteen to twenty percent.

The big picture issue here is whether Congress ought to be making the telecom industry tax collectors at all. For business, telecom taxes are pure overhead that crimp a company's ability to create jobs and pay for expansion. And if the company is profitable, it is still going to pay taxes on the profits. From an economic development perspective, telecom taxes are a drag on jobs development and business growth. Design Nine, as an example, gets phone bills with as much as 30% of the charges just taxes of various kinds, from local, state, and the Federal government.

Given the weak state of the economy right now, let's hope Congress does the right thing and extends the ban on Internet taxes for services.