I found this article by Ernesto over at TorrentFreak ("Decluttering The Tubes, Solutions to the BitTorrent â€œProblemâ€?") to be very interesting and open-minded, but his readers are really taking him to task for it. In the piece, Ernesto outlines the upsides and downsides of 6 possible ISP responses to the "BitTorrent Problem," which has been in the news a great deal lately. (These models were apparently suggested to Ernesto by Art Reisman, who is chief technical officer at APConnections):
1) Ask for voluntary cooperation.
2) Keep connections within the providers network.
3) Usage based quotas.
4) Limit the total connections allowed at one time per user.
5) Build out networks to handle the increased load and pass the cost onto the consumer.
6) Cancel the service of users who abuse their privileges. There have been reports of providers doing this already.
[Again, see full article for explanation of strengths and weaknesses of each.]
I think many of these solutions sound quite constructive and could possibly be used in some combination to alleviate network congestions problems. But the reader response over at TorrentFreak, which obvious skews towards the heavy BitTorrent user, is perhaps all too predictable: Just give us more capacity!
In fact, a great number of the comments following the article run along these lines:
"ISPâ€™s should just quit bitchinâ€™ and make their networks bigger, better, and faster. Problem solved."
Um, yeah dude, that would be wonderful, but do you think broadband networks just fall from the sky like manna from heaven? Network expansion clearly should, and will be, part of the solution. But these investments cost money. Lots of money. And it is a business, not a charity. Carriers will need to recoup their significant fixed infrastructure investments. And that can't be done without balancing costs and managing networks to some degree.
Just take a look at the grief that Verizon is getting from Wall Street analysts right now for their multi-billion dollar FIOS roll-out. Many smart people who monitor this industry have deep reservations about the ability of Verizon to pull this gambit off. For example, here's a passage I found in a recent private newsletter from Craig Moffett of Bernstein Research, one of the very best communications and media analysts in the business:
Verizon's own estimated average cost of $817 per [home passed] and $718 per connection â€“ and assuming a conservative $200 acquisition cost per customer acquired â€“ yields a total cost per connected customer of approximately $4,000, assuming a 40% connection rate. If penetration were instead to be 30% of homes passed, cost would approximate $5,000 per subscriber. Even at penetration of 50% of homes passed (that is, penetration roughly equal to that of the U.S. cable industry today, before the entry of the TelCos), cost per home would be $3,400 per connected home.
Those are astonishing numbers, folks, and if you happen to be a Verizon shareholder, you are probably left wondering how good of an investment the company is today with an financial nightmare like that in the works. Don't get me wrong. I absolutely want Verizon to succeed. I am lucky enough to live in one of the very first neighborhoods that was ever wired for FIOS. I am here to tell you that it is wonderful service. But my neighborhood is also a good example of what the problem is here: I'm one of the very few people who subscribes. Why? Because like much of the rest of America, there are plenty of folks in my neighborhood who remain perfectly happy with cable-provided broadband, or even slower telco-provided DSL connections that start at a much lower rate.
A dreamer could say, "Well, just drop fiber to DSL prices!" Sure, and GM should start selling Cadillacs at Chevy prices. It's just not realistic. If we want better, faster networks, we need to come up with new business models that will allow them to flourish. And existing networks will need to be better managed to deal with legitimate traffic management issues.
Luckily, as Matt Sherman points out, we already have some interesting competition underway in terms of broadband models in this country. And the models outlined in the TorrentFreak story sound like a pretty good place to start a more elaborate discussion about alternative models. Sadly, it's clear that it's a discussion many BitTorrent users aren't willing to engage in.