Web Informant #308, 8 December 2003

A new way to do wireless networks

 

http://strom.com/awards/308.html

 

Comdex wasn't any thrill this year: call it the demi-Comdex, the mini-me Comdex. But one meeting made the entire trip almost worthwhile. I spoke to a couple of guys who are doing a fixed-point wireless ISP. The concept isn't new, but their timing may be perfect.

 

And these guys, who are from a company called Slice Networks in Nebraska, may actually be ahead of a very big development: last week Intel, IBM and AT&T endorsed their concept with a new company called Cometa Networks that will roll out a new nationwide wireless network.

 

What is so new about it? Well, as I said, fixed-point wireless has been around a while. Metricom is perhaps the most notable example of a company who tried to pull it off but failed, exiting the stage several hundred of millions in the hole. The new take on wireless tries to do something different:

 

First off, these approaches from Slice and Cometa are making use of existing telecommunications infrastructure. The biggest challenge for Metricom was getting access to towers to build out their network. Granted, their towers weren't anything like the cellular industry needed, but you still have to put your antennas around the countryside and that takes people on the ground and lawyers ready to be deployed at a moments' notice. So if you can leverage existing points of presence that are already wired to the Internet and just add the radio gear, you are a leg up and several months' or years' ahead of the game. Of course, these points have to be near where your customers need the radio coverage, too.

 

Second, we aren't talking about needing any new gear on the client side either. Where Metricom fell short was in using custom radio receivers for each client. These new approaches make use of the most common wireless adapters out there -- the 802.11b wireless network standard. And what is nice about this standard is that the interoperability wars have already been fought and won and it pretty much doesn’t matter whose gear you use: everything (usually) works together.

 

What makes the 802.11b (also known as WiFi) play more compelling is that Intel, Apple, IBM, Toshiba, and Dell are all planning (or doing it already) on incorporating wireless networking clients in their laptops, so there is literally nothing that users have to do to take advantage of this. Microsoft is also helping matters by making Windows XP the most wireless-friendly operating system out there. (Some would argue that XP is too friendly, making it easier for users to tap into wireless networks that they shouldn't be using, but I won't take that up here.)

 

Finally, you want to leverage the existing billing and customer support infrastructure of a standard ISP or phone company to make the whole ball of wax easier. This is what the Slice Networks guys are doing, and what some other forward-thinking ISPs are planning in this area, and why AT&T is part of the Cometa play. One of the reasons that DSL has been such a boondoggle has been ginning up the billing apparatus to keep track of large numbers of people spending small sums of money on a monthly basis: the phone companies are very, very good at this, and only some ISPs are just now getting the hang of it.

 

Okay, so what's the big deal? The added wrinkle to what Slice and the bigger players are doing in fixed-point wireless is to extend the range of the 802.11b radios. Typically, these devices only operate within a few hundred feet of their access points, and that is being generous. The catch is being able to boost the signal strength to cover miles of territory, so that fewer antennas and access points are needed. In addition to boosting the signals, you need additional electronics to ensure that more than a few users can connect and make use of the access points that were originally designed to just support small workgroups. Fortunately, the technology to make this all happen is now readily available, and is coming into wider deployment. It is an exciting time. My colleague Dave Molta from my former alma mater Network Computing recently wrote a couple of articles about these technologies:

http://www.nwc.com/1324/1324f2.html  (wireless bridging)

http://www.nwc.com/1324/1324f1.html (point to point wireless)

 

So what we have here is a transformation taking place. The original idea with 802.11 networks was to provide mobility, as users roamed around their campuses with their laptops but were able to stay connected and get their work done. The new wide-ranging 802.11 application is to provide connectivity still, but to users who aren't moving around necessarily: they just are either too cheap, or too lazy to deal with getting higher-speed connections and still want to use the same gear but from a stationary spot, located inconveniently far away from any broadband supplier. Even though the growth in cable modems has been tremendous over the past 18 months, and DSL is still gaining converts despite the financial debacles of the major suppliers, there are still many locations around the country that will never be wired for broadband access. The new fixed-point wireless changes that equation substantially.

 

The guys I met with from Slice told me that they were profitable from the first month forward with their setup: the cost of goods to extend their existing points of presence into the wireless arena were measured in just a couple of thousand dollars, and it doesn't take many customers to pay back that investment quickly. Granted, Cometa has a bigger nut to crack, but I think both are on to something. Maybe this time fixed-point wireless will be a hit.

 

To subscribe, send a blank email to:

                     informant-subscribe@pez.oreillynet.com

To unsubscribe, send a blank email to:

                     informant-unsubscribe@pez.oreillynet.com

 

Entire contents copyright 2002 by David Strom, Inc. 

David Strom, dstrom@cmp.com, +1 (516) 562-7151

Port Washington NY 11050

Web Informant is (r) registered trademark with the

U.S. Patent and Trademark Office. 

ISSN #1524-6353 registered with U.S. Library of Congress