I Haven't Been Giving Cable Companies Enough Credit

Earlier this week I attended my first Broadband Breakfast Club, an innovative series of breakfast meetings in DC aimed at providing a venue for the various stakeholders in broadband policy debates to come together and discuss/debate issues in a more casual, less structured setting.

While eating I had the great fortune to sit next to Kyle McSlarrow, president and CEO of the National Cable and Telecommunications Association, the principal trade association for the cable industry.

Now, I'll be the first to admit that I tend to be hard on the cable industry. Their definition of "broadband" doesn't include enough upload capacity and I know firsthand how unsteady connectivity in both directions can be based on the usage of my neighbors. Plus I've had doubts about whether DOCSIS 3.0 could really ever compete with full fiber networks.

But as Kyle and I chatted he started lamenting how much capacity cable networks have to set aside to carry analog TV channels. So I asked him how much capacity could they deliver if it weren't for TV? The answer: 5Gbps.

Needless to say, my eyes were opened. For the longest time I've been assuming that full fiber networks are the only way to achieve first 100Mbps and then 1Gbps to the home, but now I know it's at least possible that we could have a day where there are two pipes into the home offering that much connectivity.

I've long thought that eventually we'll evolve into a full fiber monopoly as that will be the only infrastructure capable of delivering enough capacity for the next generation of big bandwidth apps. And while I still think that's ultimately true in the long run as the demand for bandwidth will never stop growing and copper-based systems have physical limitations that full fiber networks don't, we may be a lot further away from that then I'd thought.

Also noteworthy was a discussion I had afterwards with someone who will remain nameless in case he'd get in trouble for his candidness. He quite frankly stated that the reason why current cable systems have such low upstream speeds was because when the cable companies were designing their networks they didn't foresee the explosion of P2P and Web 2.0; they were thinking about these pipes primarily as avenues for entertainment to move downstream. The problem they face now is that structurally it's very difficult, if not impossible, to significantly increase upload speeds.

The exciting thing, though, is that the introduction of DOCSIS 3.0 remedies this structural limitation. While I still have reservations about how shared these networks are in relation to the performance they'll be able to deliver, at least cable companies deploying DOCSIS 3.0 will have the freedom to increase upstream capacity to meet the needs of their customers.

In truth these revelations have not shaken me off my line that the endgame for the Internet is a Full Fiber Nation, but they have opened my eyes to the fact that cable companies will continue to play a significant role in enabling our broadband economy for the foreseeable future. And that one important consideration for our formulation of a national broadband policy is figuring out how we can free up more of the capacity locked down in these networks to delivering analog TV so that we can continue to realize competition in the broadband marketplace even as speeds increase to the gigabit level.

Also, for anyone who's based in or will be in DC Dec. 9th, I highly recommend you check out the next installment of the Broadband Breakfast Club. I'll be a speaker this time around joining esteemed colleagues talking about broadband applications and what can be done to turn the rhetoric of broadband's transformative powers into reality. You can register here.