In this morning’s newspaper, there was a lengthy article by Anick Jesdanun of the Associated Press entitled, “Wireless Worries.” It starts off describing Lompoc, California’s muni Wi-Fi system and provides some interesting reading. I wrote about Lompoc’s muni Wi-Fi efforts more than two years ago, mentioning that I thought the business model was flawed, it would run into problems with interference and that its police department had been promised full mobility access in their patrol cars. My conclusion was that the Lompoc system, like many others, would not survive. Here is what the article has to say in part.

The $3 million budget for the project that was launched a year ago in this city of 42,000 has only a few hundred subscribers, not the 4,000 needed to start repaying the loans from the city’s utility fund. In the course of the past year, they have discovered that indoor coverage is basically non-existent, especially inside houses built of stucco (most of them) since stucco walls are constructed with a wire mesh embedded in them. They have also discovered that customers are unhappy with the data speeds, and that the system could cost more because more access points will have to be added.

My favorite quote in the story is by Anthony Townsend, research director at the Institute for the Future, a think tank. He said, “They are (muni Wi-Fi systems) the monorails of this decade: the wrong technology totally overpromised and completely undelivered.” Another quote worth repeating is, “I will be surprised if the majority of these are successful and they do not prove to be drains on taxpayers’ money. The government is getting into hotly contested services,” a statement by Michael Balhoff, former telecom equity analyst with LeggMason, Inc.

Mr. DeWees, Mayor of Lompoc, is now admitting that the city may have to pull the plug on the system but will try one more push and recently reduced the price by $9 per month to $16 per month. One customer said in the article, “There’s an antenna literally at the curb of my house, but when I’ve tried to log on, it cuts in and out.” And this is typical of most customer responses, at least in Lompoc.

The article goes on to address the issue of the number of access points: “Many cities and vendors underestimated the number of wireless antennas needed. MobilePro Corp’s Kite Networks wound up tripling the access points in Tempe, Ariz., adding roughly $1 million, or more than doubling the costs.” Kite’s CEO pointed out that, “The industry is really in its infancy, and what works on paper doesn’t work that same way once you get into the real world.”

The article ends with a quote from the wireless consultant who worked with Lompoc on this project: “Even if the whole network were to be written off tomorrow, prices dropped (on other services) and quality of service went up. That’s the way a lot of cities look at it. They don’t look at business profits and losses. They see it as a driver for quality of life.”

The lesson here for consultants, I gather, is that if a system you helped develop and design fails, it has still achieved its goal by driving DSL and cable services lower, and the quality of life in these cities goes up because more people are connected. Talk about making lemonade out of lemons!

Imagine my surprise at all of this! Muni Wi-Fi systems that don’t provide coverage indoors, systems that require many more access points (and backhaul) than planned and to get indoor coverage, many systems are having to provide or sell customers a mini-repeater that can be located near a window and repeat the signal back into the house.

I keep hearing reports that there are a few, mostly rural systems that are up and working and providing good service. But these are in areas where, at the moment, there is no competition from DSL and cable providers. Where there is already DSL and cable, these muni Wi-Fi systems are, as I mentioned in a recent blog posting, becoming feeder systems for the DSL, cable and even wide-area wireless companies.

From a customer’s perspective, muni Wi-Fi has proven to be a disappointment in a number of ways. First, it is not as reliable as DSL and cable, and then there is the issue of having to share bandwidth, which means connection speeds vary throughout the day, depending on how many customers are connected to an access point or cluster of access points. Today’s data speeds might be fast enough for Web access and email, but they are certainly not up to the task of handling video services.

Last year was “The year of muni Wi-Fi” with lots of hype and lots of cities announcing plans for muni Wi-Fi services. San Francisco, which has been working on a plan for more than a year, has still not decided if it will install its own system or permit the bidders to build it out. Meanwhile, the intersection I monitor each time I am in San Francisco gets worse. Three years ago when I first stood on the street corner and indentified access points, I found sixteen. Twelve, as I remember, were not encrypted and could be accessed by anyone. I have checked this corner over time and a few months ago I counted 42 access points with strong signals at the corner, and there were still more than ten that provided full and open access. I wonder how muni Wi-Fi will work at that corner?

These systems are all using a local-area technology in an attempt to establish a wide-area system. Each access point mounted on a lamp post provides a circle of coverage of between 300 and 500 feet, and if you happen to live in an apartment above about the third floor, the signal is very poor or non-existent. Coverage inside buildings is also a problem, and I still don’t see many people sitting on park benches accessing the Internet. As I travel around the United States, I am also finding more places where hotspots have been set up and are providing free access. The last hotel I stayed in had free Wi-Fi service and a free access computer in the lobby. The Dayton, Ohio airport has computers near most of its gates that provide free access to the Internet using high-speed connections. Of course, my wide-area broadband connection is almost always available to me and with EV-DO Rev A installed, it flies!

This year will be interesting for muni Wi-Fi services. Some have been around long enough that the problems I predicted are beginning to rear their heads. Last year, common practice among muni Wi-Fi vendors was to design systems that used thirty access points per mile for coverage, with a wired or wireless backhaul connection required for every third or fourth access point (the systems are based on mesh technology so one access point becomes the backhaul feeder for three or more other access points). If you look at the systems being planned today, you will see that the number of access points per square mile has risen to sixty or more, which also means that the number of access points that need to be connected to the backhaul system more than doubles. While this new configuration will provide better speeds per customer (there should be fewer customers sharing an access point), it certainly drives up the costs of these systems.

Yet the surge to build out muni Wi-Fi systems continues. According to the AP article, as of March 2007, there were 180 muni Wi-Fi systems in operation in the United States. Many are in smaller cities and towns, but there are also a number in major metro areas. There seem to be enough of these systems to be able to judge their success, or lack thereof, this year, and with more coming online, we should start seeing more data points.

The next issue is that most of these systems are being built using 802.11b, the slowest of the Wi-Fi standards, although a few are using a combination of b and g. The introduction of 802.11n, which should be finalized early next year, may be tempting for some systems. N provides a lot faster connection and a range increase—at an increased cost. Perhaps muni Wi-Fi vendors will go back to cities they have deployed and seek upgrade fees to n in order to cover the cities better and to promise faster connections to the Internet. But if they stick with their existing backhaul, increasing the airlink speed won’t help their customers at all. The choke point, as with most data systems, is not the over-the-air data speeds but the speed of the connection from the system back to the Internet. I am always amused when I am told by someone that they upgraded their b-only system to g for more speed. The first question I ask them is if they also upgraded their connection to the Internet. They usually look at me and ask, “Why would I do that?” Then I explain to them that if they are, for example, using DSL with 1.5 Mbps down and 786 up to connect to the Internet, their upload and download speeds will only be as fast as their DSL service, not as fast as the wireless link even if they are the only person using the g system. Many people don’t understand this concept; it is foreign to them, even those who work with Wi-Fi fail to grasp the concept that every system has a choke point and the data will flow only as fast as that choke point will permit.

Back to muni Wi-Fi, I will repeat my predictions. Muni Wi-Fi will turn out to be a short-lived but highly publicized way of bringing high-speed broadband to the masses. These networks will remain temperamental, requiring more access points and more backhaul over time. Interference will continue to plague these systems, either because of more access points being deployed in homes and offices (including many cable TV companies that now routinely ship an access point built into their cable modem), or of a more deliberate nature by others who do not want the muni systems to succeed or who feel that their own systems have been degraded by the addition of muni Wi-Fi.

Further, as demand for more bandwidth continues to grow for IPTV, streaming video and other high-tonnage uses, these networks will collapse under the strain. At some point, cities will realize that providing broadband wireless services to all of their population presents some very tough technology challenges, some of which probably cannot be overcome quickly and easily. But until then, the 180 cities cited in the Associated Press article will be joined by others that still believe the hype that Wi-Fi broadband access on a citywide basis is a hot issue and something they should do for the good of their citizens.

In Lompoc, among other places, the famous line, “If we build it they will come,” has not been the case. Lompoc said it needs 4,000 customers, which is only 10% of its population, to break even. After a year in operation, it has an uptake of perhaps 1% and I am willing to bet the churn rate is astronomical!

The companies that sell this equipment to the cities or that take on the building of a muni Wi-Fi system still make it sound easy, fast and affordable. They say the systems will pay for themselves quickly, free access at slow speeds will entice customers and higher speeds for a monthly fee will pay for the network, ongoing maintenance and backhaul and connection charges. In short, cities are being misled.

Wi-Fi is a local-area technology that works great indoors as a hotspot at a café, inside offices and homes. It is not a wide-area technology and there is no way to turn it into one. You would think that stories such as the one that appeared in today’s newspaper would have a sobering effect on other cities’ plans, but with so many “interested parties” pushing cities for rights to the muni Wi-Fi gold mine, my bet is it won’t change a thing and is destined to be repeated time and time again over the next two years—or perhaps embarrassed city councils will simply turn off the lights quietly and slink away into the night.

Andrew M. Seybold