Municipal Wi-Fi "experts" have egg on their faces

Wireless Fidelity or Wi-Fi was touted as the universal solution for municipalities by many who proclaimed to know everything about everything when it came to network infrastructures. In many cases, cities did not turn to seasoned professionals who would have told them that they needed something more substantial or that the network topologies that were being cited were inadequate. Instead, they bought off on the hype of a new wonder technology and its evangelists.

The realization that the technology bit off more than it could chew is nothing new. In a recent article in Fast Company, the comments reflect a new realization that Wi-Fi isn't a universal solution to citywide networks:

And yet, the leading muni-fi builder, which has committed millions to building city Wi-Fi networks, still hasn't nailed down the formula for profitability.

"It's an evolving process," says Cole Reinwand, EarthLink vice president of product strategy and marketing. "That's what we're trying to get a handle on right now."

The Wi-Fi tab For one thing, the total upfront cost is still an unknown. The number of nodes, or antennas, that EarthLink and hardware maker Tropos planned on using in Philadelphia had to be doubled to get better reception, according to Wi-Fi expert Glenn Fleishman of Wi-Fi Net News. At $2,000 or $3,000 per node, switching from some 20 nodes per square mile to almost 50 quickly runs up the bill.

In designing networks, one thing that you find is that it's not a cookie-cutter approach, and what works in Philadelphia may not work in San Francisco or Chicago.

In emerging technologies, many people think that there is some standard recipe for success. A long-time friend and colleague, Mike Carioscio, who is CTO of Pepsi, used to say that many look for “some magic bullet to solve all network designs.” There aren't any.

There are many variables that you must consider, as well as the environment in which you are putting up nodes. More nodes translate to higher implementation costs. All of a sudden, a bargain approach now requires a huge multi-million dollar cash infusion to make it work. That is happening in some places.

When in doubt, blame someone else

At Duke University there were complaints about people not being able to use their iPhones and they somehow crashed some Cisco networks. Hmm, how many experts put that wireless network together? And what about the experts commenting on the real problem? From an expert blog:

When Wi-Fi access points began issuing the equivalent of busy signals, Duke network staff traced the problem to misbehaving iPhones that were flooding the routers with bogus Internet address requests - at the rate of up to 18,000 per second.

Unfortunately, or fortunately, the author had comments from readers that had a more realistic perspective. This one was one of the best - and most accurate:

Several iPhones - perhaps as few as two - flooded campus routers with IP address requests, at the rate of up to 18,000 per second.

I've been working as a Cisco engineer for a long time, and never have I heard anything like this!

First of all, Cisco wireless access points/routers can't handle 18,000 requests a second.

Second of all, iPhones must be using some powerful processors to be requesting DHCP address 9,000 times a second.

All of the misinformation - blame and pseudo-expert hype in this article - proves my point again that there are a lot of people writing and commenting about this stuff that really should not hold themselves out as an expert.

From another article, the iPhone is exonerated:

A problem with Duke University's wireless network caused outages at the school, officials said Friday, exonerating the initial suspect, Apple Inc.'s new iPhone.

"A particular set of conditions made the Duke wireless network experience some minor and temporary disruptions in service," Duke spokeswoman Tracy Futhey said in a written statement posted on the university's website. "Those conditions involve our deployment of a very large Cisco-based wireless network that supports multiple network protocols."

The question then becomes, what did the design experts do wrong at Duke? The bottom line is that many networks have not been thoroughly tested or set up to be very secure.

Wireless networks can be secure, but you need to know what you are doing. And as for adequate bandwidth...

Garden hoses don't put out fires

Just as you would not expect firefighters to use garden hoses to put out fires, you cannot expect network infrastructures to deliver huge amounts of bandwidth if you are using a wireless network that was never designed to be a fire-hose of bandwidth.

Maybe if we explain it this way, more people will “get it right” instead of “getting burned” by inadequate network designs. There are other significant wireless alternatives, like WiMAX, but we will save those for another time. There are also other carriers getting into the municipal solution business.

Why? More municipalities are looking for wireless solutions and other alternate network infrastructures because they are concerned about economic development. They are reaching beyond the traditional incumbent's approach and timeframe because they cannot wait to get something better in place. Some have locked in on wireless as the solution that is needed.

Now that AT&T is getting into wireless solutions for cities, some cities believe that if they are supplying the answer, it must now be a safe bet. If nothing else, AT&T's entry into the wireless solutions market provides more credibility to the technology as a potential viable solution.

However, there are no safe bets in this industry anymore. No one has a corner on talent in the industry and even the carriers can get it wrong at times. AT&T came in late into the market compared to others, but that is not surprising.

As I commented on a column in MuniWireless last September:

Incumbents are not risk takers or innovators and will wait long before committing any resources until a market is labeled a "sure thing."

So AT&T followed the old adage, “If you can't beat `em, join `em” and pursued a wireless solution for Springfield, Ill. as well as bidding on Chicago's wireless initiative. They are courting other cities as well.

But with large cities, you will need a combination of both wireless nodes and fiber optic infrastructure to handle the traffic. Many evangelists for Wi-Fi don't have the broad background that you need in order to design and implement real mission- critical networks that may need a combination of transmission media.

For years, I started network management courses with:

There are no experts in this industry, the best you can be are good students - always learning.

Unfortunately, many who proclaim to be wireless experts are not.

CARLINI-ISM: You get what you pay for. When you pay peanuts, you get monkeys.

Recent articles by James Carlini

• James Carlini: How good (or bad) is your local emergency plan?

• James Carlini: Product dumping, labor dumping - it's all the same

• James Carlini: Green is the color of speed in ranking broadband infrastructure

• James Carlini: A disgusting video: U.S. companies sell out Americans

• James Carlini: No network infrastructure, no Olympics, no nothing