Saturday, July 26, 2008

Home Networking, Verizon, and Lafayette

Food For Thought Department
[What follows is lengthy and starts out with arcana but I think the implications are significant—perhaps especially for Lafayette. I ask that you stay with me...]


According to TelephonyOnline Verizon is radically upgrading the gateways it installs in homes served by the fiber-to-the home-based FIOS service. —FIOS customers can buy a triple play internet/cable/phone package from Verizon based on technology that is very similar to that being constructed in Lafayette by the community's Lafayette Utility System.

The new in-home devices have a number of interesting characteristics; they will:
  1. bump "speeds over coaxial cable in the home from 75 Mb/s to 175 Mb/s"
  2. "have double the processing power" compared to the current gateways
  3. allow "users to create up to four separate wireless networks, each with different security settings"
  4. allow "remote Verizon technician management"
Understand that an upgrade like this is costly. Customer Premise Equipment (CPE) is costly. Putting a piece of relatively pricey equipment in every home (on top of the set-top boxes you've installed for video and any VOIP equipment) really adds up. CPE is where every company tries to pinch pennies and extend the life of its equipment. So upgrades are rare. And they are never done without a damn good reason.

So why would Verizon invest in new hardware with the hopes of using the new capacities in “the next three to four years?”

My best guess: to ride the wave of big bandwidth in the home...Big bandwidth inside the home has recently emerged as an issue. (I've just recently caught on. See my recent post, FTTD (Fiber To The Desk, for some background musing on how really big in-home transfer might be accomplished. What Verizon is doing validates the idea but is pretty small potatoes compared to what is coming. Don't miss the comments--good stuff there.) Verizon clearly thinks that its current model which provides 75 mb/s will prove inadequate for in home use in the next 3-4 years. Pause to let that soak in please: the next 3-4 years. Tomorrow.

That is a near-future time frame. Nobody spends the amount of money that Verizon will spend even gradually moving over to new equipment without a very compelling plan to make back their investment. And Verizon knew what it wanted in these boxes. These are not off-the-shelf pieces; they've been designed to Verizon's specs and the company has contracted two independent providers that meet those specs in order to assure itself of supply.

So the difference between the previous standard and the new equipment should strongly hint at what Verizon thinks folks will do that makes the upgrade pay out. Let's unwrap those specs looking for clues:

The Analysis
—Faster speed, from 75 to 175 mb/s, means that Verizon is expecting a lot of internal traffic on home networks. That is lot-—especially since Verizon won't offer you more than 50 megs of connectivity to the internet itself so it's not video downloads they're trying to accommodate (of course not ;-) ).

So for what do you need massive amounts of in-home networking speed? Take a gander at the processing power for a partial answer.

—"Doubling the processing power"--if you dig around a bit (1,2) you'll see that that phrase refers to moving from a 32 bit chip architecture to a dual core 64 bit chip. That's the way my fancy laptop is built. That's real processing power even if the clock speed turns out be a bit lower. It allows the onboard computer to coordinate more in-home devices. Most obviously multiple set-top boxes for the cable video service are in the mix; it takes a lot of bandwidth to push HDTV around especially if one or more set top boxes is acting like a DVR/video server and pushing video out to secondary screens. In fact the "doubling" phrase clearly understates the added computational capacity. On top of chip architecture the gateways can serve out eight (8!) Quality of Service (QOS) controlled channels. At a minimum that means that Verizon can push 8 separate protected streams to multiple TVs. But eight seems like more than homes really need. Of course there are Xboxes, and Wiis, and Apple TVs and the like in addition to a raft of reasons that users or companies selling to users might want a protected stream...So eight makes sense. . . If (and only if) you are planning to do something beyond video.

Ok, its faster and more powerful...

—What's with that feature: allow "users to create up to four separate wireless networks, each with different security settings?" Well, that allows you to set up a buncha different networks, some with QOS, some without, some that are slow, some that are public.....hmmn, whats with that? Again, it seems like overkill for current services. Wireless is hard to maintain for the QOS that video requries. They are probably wisely sticking with wires (coax) for that function given the eight protected streams on the wired side. By any measure the capacity for 12 separated streams is pretty astonishing.

Faster, more powerful, many separate streams, eh?...

—Finally: allow "remote Verizon technician management." That means that Verizon can modify the thing from their headquarters. That alone isn't too new—most "modems" can be upgraded by the company or at least reset to clear glitches and given the clearences it needs to access the ISP's network. But in this context "remote management" surely means the ability for Verizon to enable and assign all those streams and to install some management software or special access codes on unit. And, sell that capacity to third parties who would like to use the in-home network that Verizon's fancy new gateway creates.

Faster, more powerful, many separate streams, that can be controlled by the network owner...extending its control of the last mile into your rooms.

Chew on that for awhile. I did.

The Conclusion
Some folks might think all these bells and whistles are just over engineering. I can't believe that a traditional telco like Verizon, one that is already straining its financial capacity to pay for a fiber build, is investing that kind of cash unless they really think this amount of capacity will be valuable to them within 4 years and pay for itself rapidly at that point.

My guess is that Verizon wants to control your home network and all the things that you are shortly going to want to run on it. Things which you might want today if only it weren't so hard and costly to get the service up and running.

What Verizon wants to sell you directly is only the base. Video and video serving is likely only the beginning from Verizon's point of view. The corporation has two profit centers currently: data and wireless. (Video shows promise but isn't there today. Old style telephone lines are shrinking.) Convincing you to buy more data capacity and their wireless service is a proven cash cow. Wireless's Achilles heel remains coverage and the most persistant irritation. In-house and in-building coverage is a big problem and one that is hard and expensive to solve by popping up more cell towers. The emerging solution is to use "femtocells" —to set up a small base station inside the building that is hooked up to a wired network and provides a mini "tower" that dramatically improves service. An in-home gateway like the ones described could help service and manage the bandwidth and protocols necessary to easily deploy this service to those that need it. And potentially radically reduce expensive customer turn over.

But, as popular as video and wireless retention has to be with the accountants who like old services and guaranteed returns, the real goal is likely broader: providing a platform for other, secure, protected services. Services which people can be sold but for which each provider currently has to figure out how to provison. A truly capable gateway like the ones that are described would let a lot of service providers play without installing their own in-home network and/or controller device.

All Verizon would want is, say, 20% off the top.

And providers would probably find that cheap compared to installing their own network.

Here's an unordered list of things which would be much more commercially viable if the infrastructure/platform were already installed in the home and could be activated and managed remotely:
  • Gaming can eat local bandwidth too.
  • Virtual Private Networks (VPN).
  • Video telephony and intercomms.
  • A local high school sports "network's" video stream and pay per view.
  • National professional and college sports "channel" versions of streaming video and downloads direct to your DVR.
  • Sophisticated security networks and security cameras.
  • "Telepresence" and other video conferencing/telephony.
  • Allowing all your electrical devices AC, refrigerator, hot water, etc to communicate and lower energy costs.
  • "Smart home" sensor and activity networks.
  • "Satellite radio" channels.
  • A myriad of music rental services could play directly through your connected sound system.
  • And many more.....add your own in the comments
Many of these "long-tail" sorts of uses will be "gotta have it" for some subset of users. The 2 dollar USL sports network will lock in users who will spend the next 200 monthly dollars on Verizon. Suppose only 1% of users has gotta have each of the above functions. That's 11% of the market right there. Locked into your company from the start. I think Verizon could make a pretty penny by controlling the gateway device that made such home functions easy to buy, install and provide.

I think any network could.

The Take Home
Understand that the current incumbents know very well that the only thing that keeps them from becoming cheap, commodified transporters of other people's expensive bits is their monopoly-based control of the last mile architecture. If we had six connections to the outside world all networks would be running cheaply and competing on how fast and reliably they could provide us with bits. (But that ain't in the cards...which is why wise communities will follow Lafayette's lead.) The incumbent's stranglehold on the last mile is crucial to their profit profile. With it they are Gods of the network age. Without it they are the guys who sweep the roads and fill in potholes—and will be paid appropriately. They'd rather be Gods. That last mile control is the key to being invited in to control the network in your home too and the key that will give them huge new sources of revenue by controlling the toolbooth that they hope that new gateway will become.

It's a pretty damn good plan.

Lafayette
No Lafayette Pro Fiber blog post would be complete with a comment on the local implications. IMHO this is another place where Lafayette could lead the way.

Verizon is poised to extend its fiber-based advantage into the home by controlling access to big bandwidth inside the home and by easing the entry of services that critically depend upon accessing a robust home network. In some places the cable company has already partnered with security firms to provide robust networking that rides on the coax installed for cable. Other incumbents surely will see the writing on the wall; they'll have to follow suite or watch companies like Verizon generate the revenues that will enable them to become more and more dominant. Verizon has the big bandwidth advantage in fiber. But that advantage is purely theoretical until the public can see that the more capable network can provide not only "more of the same" but actually "different and better" services. The gateway can be the key that unlocks that potential.

A gateway or something similar can do the same for Lafayette's network.

I've been hearing a lot of background buzz lately about trying to encourage tech development in and for Lafayette. Meetings in various places, varying level gurus flown in from various places to attend the meetings. Dinners in posh private homes. Talk about establishing an "x-prize" for Lafayette. An attempt to organize a meeting for developers. Desultory attempts at secrecy. (My list is surely incomplete.) The usual influentials' names are bandied about. You know, the works. No one knows whether any of it will come to fruition. But the point is that Lafayette is beginning to wake up to the fact that it will be well served to actually do something to encourage development. A "build it and they will come" attitude only works in the movies. In the real world if you want something to happen you've got to do something special to encourage it. Building LITE and LUSFiber and ramping up LCG's example are great starts but they won't, alone, be enough to make Lafayette the mecca many of us would like to see it become.

So far most of the Lafayette discussion on this topic has been couched in terms of somehow convincing (or bribing) developers to make us something special. As much as I like the idea—and hope it succeeds—I think we'd have better chance at success if we instead tried to do something special ourselves.

Like Verizon is evidently doing.

The heart of Verizon's apparent plan is to make it possible, even easy, for developers to do something great and different. They are poised to eliminate the barriers to "getting things done" by providing the platform over which these things can be accomplished. Verizon lays out all the tools on the table (albeit tools that lock you into their network) and will surely even handle billing for you. But they won't pay you. Instead they'll charge you...and your customers. Frankly, that's a better way. Opening the door is always a better plan than subsidizing the battering ram!

The right box in the house could do for Lafayette what Verizon's gateway is poised to do in the homes of it FIOS users. But Lafayette's could be based on ethernet and open IP standards instead of the clunky cable-oriented and proprietary network hardware and protocols that serve Verizon but are unfamiliar to most developers. Lafayette could do a better job of facilitating access to the Local Area Network (LAN) that is the home than any of the competitors is willing to do.

But Lafayette could go further. It could do the same for its MAN (Metropolitan Area Network) by building in the resources that make access easy. Make available storage. Make available the kind of computational power represented by LITE and Abacus. Embed modern protocols. Pack up some servers to enable within network serving of various kinds of data (streaming video, for instance).

In short Lafayette could make its networks, both inside the home and inside the city, playgrounds for the easy, fluid kind of development that developers love.

And we might, eventually even make a few pennies off it. But quickly and surely we could make Lafayette a tech mecca, give LITE a clear purpose in the community, and make LUSFiber a roaring success.

You want to be that shining city on the hill? The path is open.

No comments: