Why is this page text-only?
 

« FCC NBS Comments: Why Not Follow Japan's Lead? | Main | Four Months And Counting... Still No Broadband Stimulated »

June 16, 2009 10:43 AM

FCC NBS Comments: Agreeing and Disagreeing with Brett Glass

Continuing my exploration of the comments to the FCC on how best to formulate a National Broadband Strategy I wanted to dive into those filed by Brett Glass.

Brett heads up LARIAT, a wireless Internet service provider in Wyoming that was "among the first, if it was not the very first, of the world's terrestrial high-speed WISPs." Brett's also been a long-time independent and vocal voice on matters of broadband policy, speaking from the perspective of a smaller network operator who's committed to bringing broadband to those who don't have it. In fact, in their service area in southeastern Wyoming only 5% of their customers have access to any form of wireline broadband.

His first major point relates to the definition of broadband: "A local cable TV plant, which delivers video, telephony, and Internet services over the same cable but separated by frequency is an example of a 'broadband' system. Note that the term 'broadband,'
properly used, does not imply that Internet is even among the services carried by the system, nor does it imply any specific performance level."

On the one hand I agree with this statement as there are applications that can utilize broadband networks without ever touching the Internet, but at the same time I can't ignore the thoughts expressed by David Isenberg and others in their comments summarized by their tagline: "It's the Internet stupid." These suggest that the FCC's primary concern re: broadband should be its relation to getting people to the Internet. Ultimately I think they're both right as I don't see how we can talk about broadband without including the Internet, and yet it's also a mistake to ignore broadband's uses beyond the Internet.

Brett goes on in his comments to stress that "data throughput is not the only criterion that should be used to evaluate the quality of broadband service. Many users who switch to LARIAT's terrestrial wireless broadband service do so not because throughput was slow but because the applications they wished to run (e.g. VoIP) could not tolerate the satellite system's high latency. For a VoIP user, an online auction bidder, or a participant in an online game, a data throughput of 200 Kbps, with reliable latency to the backbone of 75 milliseconds or less, is actually far more desirable than a 5 Mbps satellite connection with an average latency of a third of a second and a jitter (variation in latency) of almost as much."

This is an extremely salient point. The definition about broadband has to be based on more than just bandwidth.

Continuing on: "The Commission should recognize that many service providers make exaggerated speed claims, quoting the maximum raw bit rate of their equipment as the speed of the service. (This is tantamount to a used car salesman claiming that an automobile is a '120 MPH car' because this is the largest number on the speedometer.) Such claims, if they are not prohibited entirely as a deceptive marketing practice, should
certainly not be honored when the actual performance of a connection is evaluated."

Another great point. The definition of broadband shouldn't relate to speeds promised by the bandwidth that's actually delivered. And, in fact, we should be looking seriously at preventing deceptive advertising practices where network operators claim they can deliver a certain level of service but in actuality can't.

Brett's last point in this section suggests that the FCC should allow the market to decide what speeds should be offered vs. dictating them. I'm torn a bit on this one. On the one hand I agree with Brett as there are great disparities between what network operators pay for backhaul connections to the Internet, which becomes a problem if you have to deliver ultra-high-speed broadband in an area with high-cost backhaul connectivity. At the same time, I worry that if we leave this entirely up to the marketplace that we may only see small incremental improvements in capacity over time and that we could end up with a lot of money invested in lesser networks.

Because of this I'd suggest that instead of focusing on dictating what levels of service network operators have to offer that we instead define broadband based on the capacity of the access technology. So when we say we want 100Mbps symmetrical to every home, that doesn't necessarily mean a network operator would have to offer that service today if it's prohibitively expensive, but what they would have to do to be considered true broadband is invest in networks with the capacity to offer a real 100Mbps in-network to every home. For wireless this number would be different, perhaps 5Mbps or 10Mbps. In this way we can encourage the deployment of next-gen networks without dictating what services have to be delivered.

Brett goes on to make some good points about the state of wireless spectrum relative to smaller operators, namely that the spectrum auction process favors the big carriers and that smaller operators like LARIAT need more spectrum and/or the ability to add power to their signals. As I'm not well-versed in the wireless world I won't comment on these ideas at this time.

In the next section, Brett touches on the need to solve the "middle mile" problem, but not just in terms of getting more middle mile networks out there, which is where most of the discussion seems to focus, but also on making existing networks available for operators like himself to use. He shares that in his area he's not far from a number of nationwide "information superhighways" but that none of them are willing to work with him, and the local ILEC he has to use instead engages in anti-competitive price gouging.

While Brett doesn't get into many specifics about how to solve this, I like his general suggestion of first incentivize the backhaul operators to open up and work with other network operators, and then if that doesn't work require them to do so. Essentially feed them carrots to try and get them moving, but also let them know there's a stick ready to drop if they don't.

Brett's final points have to deal with everyone's favorite issue: net neutrality. Interestingly, he claims that "so-called 'network neutrality' regulation...would impact small and competitive ISPs like LARIAT far more than large incumbents, potentially destroying alternative providers and eliminating competitive options for consumers."

He then goes on to share that the FCC's treatment of this issue so far, most recently centered around the Comcast/BitTorrent case, has created so much uncertainty over what is and is not OK in terms of managing networks that it has driven away potential investors from LARIAT.

We heard a similar sentiment shared by Michael Johnston in a VidChat on this site last summer, where because the FCC told Comcast what not to do but didn't say anything about what it can be doing Jackson has had to hold off on offering a higher tier of service that they were going to include restrictions on P2P usage to keep their costs under control.

Brett goes on to suggest that the FCC "always seek to solve such controversies by first incenting and stimulating competition, and intervene further only when such efforts are unsuccessful and there is clear evidence of anticompetitive behavior and/or market failure. LARIAT further requests that any regulation of network management practices be accomplished via due process - rulemakings subject to public comment and discussion - rather than 'case by case' adjudicatory actions, so that it is clear to all, in advance, what the rules actually are."

So Brett thinks the FCC should be reactive rather than proactive in dealing with net neutrality, but also that when reacting they should have a public rulemaking rather than case-by-case adjudicatory actions.

The problem is I'm not sure if those two things fit well together. If the FCC's being reactive rather than proactive then network operators can never know "in advance what the rules actually are" until someone steps over the invisible line that triggers a reaction. I can see this leading to smaller operators having to be really hesitant with everything they do as they can't afford to fight any battles at the FCC, and larger operators operating with impunity because they can afford to fight those fights.

Also, I can't help but feel that having to go through a public rulemaking every time an operator steps over that invisible line could turn into a massive headache that will slow everything down as the FCC has to deal with an influx of public comments that they have to sort through and make sense of.

In order to manage this process more efficiently I think the FCC has to be at least a little bit proactive in working to clearly establish what's OK and what's not relative to network management. When it comes to setting those rules instead of a traditional public rulemaking I'd rather there be some body dedicated specifically to tackling these challenges that could bring interests from both sides of the table together along with an arbitrator of some sort to work through the arguments of all sides. Basically get the network operators, the public interest groups, and the applications developers all around the same table to work through what's OK and what's not, what's harmful and what's helpful.

This entity could also be proactive in studying what's happening in the marketplace on a regular (quarterly?) basis to keep up on the latest technologies and techniques, which can help keep all parties informed as to what the latest developments are and also get out ahead of issues a bit before they blow up into huge problems.

I know I'm still being overly vague in these ideas, and in general I agree with Brett's concerns, but I think we need more than a reactive public rulemaking process to effectively and efficiently deal with the complex issues surrounding net neutrality.

Del.icio.us Digg Yahoo! My Web Seed Newsvine reddit Technorati

TrackBack

TrackBack URL for this entry:
http://www.app-rising.com/cgi-bin/mt/mt-tb.cgi/1418

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)