SXSW: Wireless Heresy!

Wireless is being touted as the solution for everything from linking your laptop and phone to flushing your toilet from work. Everybody, particularly those with no engineering background, have decreed that unlicensed wireless is the communications model of the future and all other modes will perish like the dinosaurs.

Throughout SXSW that's the party line you heard—unless you attended this panel. The stark reality is simple: 8Mbps for a high-def video stream, a couple streams per home, multiplied over a couple dozen homes in a neighborhood, and now you are talking about massive bandwidth needs. It's clear that wireless won't have the capacity for our typical household needs. And that's not even considering latency issues.

I believe that while wireless will dominate local and personal networks, metropolitan networks will be a mix. Heresy! That's in opposition to nearly everything we heard this weekend.

I believe that in dense, affluent communities, the network will be built with optical fiber. Wireless will serve in sparse communities, where it makes sense to deploy antennas rather than trench fiber, possibly delivering reduced services. I also think that less affluent communities may end up being served by wireless. That's because the communication utilities won't see the revenue to justify the infrastructure investment.

I'm not happy with this conclusion. It says that the communication providers of tomorrow will be the incumbent utilities of today. Right now, our choice is either the phone company or cable company. This view says that's not likely to change. Wireless held out the hope of a rich, competitive environment. Now, I'm not so sure that will happen.


Comments have been closed for this entry.

re: SXSW: Wireless Heresy!

Great essay and great notes! Thanks for coming to the panel and writing about it.

Cable companies are pushing 8 - 15 Mbps for a high-definition video stream. All sets will be HD in a few years. Two to four sets per house is probably typical. For a 700-home subdivision, maybe 1.5 streams per home peak, that's in the vicinity of 15 gigabits/sec. That ain't gonna be served over the Internet anytime soon. Not only is that viciously fast, but you're going to have to serve it from the neighborhood edge, i.e. banks of rackmount computers (video servers) and tons of hard drives in the POP for the on-demand stuff. (Large companies who have a lot of fiber may still find operational advantages in backhauling all that traffic to a larger facility, and thereby have fewer POPs.)

I want to point out that the super high bandwidth needs are driven primarily by video, and that this sort of video stuff is already being rolled out by cable companies.

For Internet alone, a bursty few megabits/sec is still really fast and really affords some breakthrough stuff. Wi-Max may be able to go really far for that kind of stuff. (If Wi-Max can really ooly truly reliably deliver 70 megabits/sec to each home, it really will be competitive. It remains to be seen what it can actually accompolish in practice, and what level of reliability it can achieve.)

My laptop has a gigabit ethernet interface and 802.11g. The gigabit is 20X faster, and really more than that, because the 54 Mbits/sec... you don't actually get that, most of the time. The machine is capable of moving hundreds of megabits a second on large file transfers. Of course it's faster to be wired, and I don't know anyone who has a lot of machines at home, who has their desktops on wifi. So my 16-month-old laptop can already drive a good chunk of a gigabit/second. And a couple generations ahead will have no problem doing that. So the computer already has an interface that can go 15 times faster than what Wi-Max is supposed to be able to make available to me. And ten gigabit ethernet is coming. So I feel like it's not going to be enough, although future wifi standards will continue to push higher speeds as well. And this wifi stuff wil be great and it will have its place among a plethora of technologies that are used based on the tradeoffs for the particular customer base and their environment and economics.

It's all cool.

re: SXSW: Wireless Heresy!

Chip, I think you went to the wrong panel. That seems a narrow vision of wireless to me... and all the predictions are suspect (including my own). After all, no one predicted that WiFi would become the tool of choice for wireless broadband.

re: SXSW: Wireless Heresy!

Wi-Fi is kind of a strawman. I agree it isn't suitable.

Talk WiMax instead. It's still sub-gigabit—more than two orders of magnitude below the capacity necessary to serve high-def video to a neighborhood.

Doesn't mean that wireless technology isn't critically important. It is. Wi-Fi is the leading edge in the transition towards spectrum deregulation.

Nonetheless, I think we are in the midst of another tech hype bubble.