Why has the idea of metro Wi-Fi lost its appeal so quickly?
The idea was simple. People had already demonstrated their desire (and willingness to pay) for broadband Internet access at homes and offices. They were apparently falling in love with all things wireless. Wi-Fi hotspots were sprouting up, even if they were mainly confined to coffee shops, airports and hotels.
So, the argument ran, if local residents and visitors could be persuaded to pay a low fee in exchange for more comprehensive, citywide coverage, municipalities could then justify the cost of building and maintaining citywide Wi-Fi networks. Such costs were believed to be affordable. The hope was that the service could be subsidised for those on low incomes, helping to bridge the digital divide.
And so, three years ago, hardly a week passed without a new city in the US announcing exciting plans to blanket themselves with Wi-Fi connectivity. Some projects even stated that access could be provided free of charge to everyone, as long as users put up with a few minutes of advertising content. Many expected the resulting networks to be powerful enough to penetrate the city's walls - therefore emerging as competitors to established fixed broadband providers.
Politicians promised to 'illuminate' entire municipalities with high-speed wireless. Hundreds of local authorities either partnered with Internet service providers (ISPs) or went on their own to begin rolling out the necessary infrastructure.
Expectations too high
But something hasn't quite gone according to plan for the city planners. By late August 2007, evidence started to emerge that - in many cases - expectations were simply too high.
EarthLink, a US-based ISP and one of the largest installers and operators of muni Wi-Fi networks, laid off half its workforce and said it was scaling back its investments in the business, starting by pulling out of a contract to build San Francisco's network.
A few days earlier, Chicago had shelved plans to blanket its 228 square miles with Wi-Fi access, while other cities were reporting technical and financial difficulties.
"I think what's gone wrong is that the original concept was flawed from the beginning," says Stan Schatt, vice president of broadband and wireless networks with ABI Research. "The feeling was that, somehow, consumers would wind up with what amounts to a free network, and that has not proven to be the case."
Schatt says the infrastructure costs were also found to be much higher than originally anticipated. The calculation error wasn't so much financial as it was technical: while it was thought that a few outdoor access points mounted on light poles or rooftops would be enough to achieve the desired coverage and capacity density levels, the reality was that many more were needed.
A new business model
So, is it all over? Was this just another over-hyped technological dream that only looked great on paper? It seems too early to tell. The consensus among industry observers is that, with the technology working just fine, metro Wi-Fi still has a chance to succeed if its business model is adequately redefined.
"What's needed, we believe, is that the municipality essentially becomes a major tenant for an infrastructure company which is willing to put the network together, and this needs to be on a long-term basis," Schatt suggests. "That gives the infrastructure provider a guaranteed source of revenue, which gives them time to develop their consumer proposition. And it's pretty clear that such proposition is not going to be a free Wi-Fi service, but it could be low-cost Wi-Fi eventually."
The idea of having the municipalities essentially bearing the brunt of the network operational costs makes more sense in the US. Since September 11, an increasing amount of federal funds have been allocated to local governments to be invested in a range of homeland security-related projects.
Last September, a $1bn grant was made available to fund the development of public safety interoperable communications systems.
Stephen Rayment, chief technology officer and co-founder of BelAir Networks, one of the main vendors of wireless mesh equipment, confirms that most of the demand for metro Wi-Fi deployments is currently stemming from "networks that provide a mixture of applications but with a pretty strong emphasis on public safety and public works".
"There are large amounts of budgets for projects that contribute to homeland security and public safety, and we've done very well in those types of deployments," Rayment admits.
He names the Minneapolis rollout (one of over 300 muni Wi-Fi networks worldwide where BelAir equipment has been used) as an example of a profitable business model thanks to the incorporation of the public safety element. "It covers 59 square miles, it's pretty much all operational at this stage and it's a mixed-use network," he says.
"There is a service provider involved (US Internet), which has purchased the equipment and has deployed and are operating the network. They generate subscription revenue from residences and from people who visit the city and want to get on the Internet for $5.95 a day.
"Additionally, they came to a business arrangement whereby the City would guarantee a certain amount of income to US Internet so as to be able to use the network for their own public safety or public works requirements. This mixture of revenue streams gives the operator the ability to pay back the network equipment much quicker."
Wireless mesh networks are thought to be well suited for applications such as CCTV camera backhaul and automatic meter reading, with some municipalities already testing their effectiveness.
European cities are unlikely to get the same level of public funds to spend on metro Wi-Fi networks. But that doesn't mean there isn't a workable business case here, too.
"In Europe it's partly a different situation," says Schatt. "There, the cost to subscribe to Wi-Fi is still outrageously high. Municipal Wi-Fi, even if there's a cost involved for consumers, generally is going to be a lot lower than what they're paying now. UK hotels are a good example - the price they charge for broadband is just extraordinary. I think there's a real potential market in Europe for municipal Wi-Fi because it becomes much more affordable than what's currently available."
There is currently a network of hotspots in the UK that have been built by companies such as BT and The Cloud. But it remains to be seen how much real competition municipal Wi-Fi alternatives will provide when the only such networks being built at the moment are owned by BT and The Cloud.
Working in partnership with Intel, BT has already launched 12 'Wireless Cities' in the UK. The Wi-Fi networks cover the city centres of Sheffield, Nottingham, Portsmouth, Bristol, Glasgow, Birmingham, Edinburgh, Newcastle, Leeds, Liverpool, Cardiff and Westminster. They are targeting businesses, consumers and local authority workers.
Meanwhile, The Cloud, which operates over 20,000 hotspots in the UK and five other European markets, launched its first city-wide network in April 2007. Using wireless infrastructure supplied by BelAir, the network offers 95 per cent coverage of the London Mile.
Two networks, one use
If muni Wi-Fi networks have already found it tricky to attract large numbers of users, their woes are only likely to intensify once mobile phone operators begin to push more aggressive flat- rate data packages for their higher bandwidth, HSPA platforms.
However, BelAir's Rayment believes there are a couple of drawbacks that the 3.5G cellular guys will have to contend with: "Probably the most important one is capacity density. Yes, you can get several Mbit/s now with HSPA, but when you work out how many Mbit/s you can deliver to users per square mile, we deliver around 20 Mbit/s from just one of our nodes, and we may typically deploy 40 or 50 of those nodes in a square mile. If you do the math, that gives you one or two orders of magnitude higher capacity density that we can deliver compared to these new cellular solutions.
"The Square Mile in the City of London is a good example of an extremely dense area," he points out. "We've got a very high density of nodes in that area, so the Mbit/s per square mile there are extremely high.
"One of the things the cellular folks often say to you is that this Wi-Fi stuff can't do mobility, but that's not correct. We have a completely mobile network throughout the entire City of London - you can start up your laptop session or VoIP phone conversation in one location and you can go anywhere around the Square Mile while keeping that session alive the whole time.
"The other key thing is devices - it always comes down to devices and services. There's close to a billion Wi-Fi enabled devices, and they're not just laptops anymore: they're phones, cameras, Microsoft Xboxes and practically every conceivable type of device. On the other hand, there are only a few million HSPA-based devices," Rayment argues.
Muni Wi-Fi and 3.5G operators are indeed in frank competition with one another in each of the cities where they coexist. And while BelAir is right to point out that 'hotzones' are usually better for sheer speed, their cellular counterparts consistently win the battle for service availability.
Independent tests carried out across the US and Canada by specialist consultants Novarum, show that all of the ten fastest wireless broadband networks belong to the muni Wi-Fi variety, with the two fastest cellular networks coming in 12th and 13th in the ranking. However, when it came to measuring service availability, Novarum discovered that only one metro Wi-Fi rollout (St. Cloud Cyberspot) offered full coverage, compared with 15 cellular data services which hit the 100 per cent mark.
ABI forecasts suggest that we can expect more Wi-Fi cities to emerge in the near future. But the research firm is also cautious: "I think we will see continuing growth of municipal networks, but it's going to be at a moderate pace," Schatt predicts, "particularly because the world seems to be moving into a recession of sorts. I would suspect that tax revenue would go down and - if tax revenue goes down - then infrastructure projects such as these will probably be affected."