The history of telecoms is littered with ventures that failed because providers leapt on a bandwagon without analyzing the specifics of their market. The LMDS broadband wireless boom at the turn of the century was a good example, with operators investing huge sums in technology that was not suitable to the end user market economics and failing to discover whether there was even a market to address. Wi-Fi metrozones are in danger of going the same way. While the appeal of ubiquitous, low cost broadband is clear and the disruptive potential of providing an alternative mobile metro network to 3G is appealing, in reality metrozone projects are still desperately searching for a working model and for a viable technology. Wi-Fi alone will not support a genuine alternative to cellular 3G, and the commercial and municipal providers need to decide exactly what their motivation is, and whether there is even a demand at all in their particular area. This is not just a discussion for public authorities ‘ even the much reported Google advertising-driven metrozone model is non-viable unless some more robust technology choices and some proper market targeting are available.
On the demand side, why on earth are the cities of New York and Paris talking about Wi-Fi clouds? These are two of the best equipped metropolises in the world with Wi-Fi access, in the New York case often at low cost. By all means offer subsidized broadband to poor boroughs, but the recent presentations by these cities’ authorities focused on the affluent downtown areas. This makes a nonsense of the most popular justification for municipally controlled metrozones, bridging the digital divide, and yet offers no clear commercial model either, since there will already be so much competition. Another common reason given is that a publicly owned network will support greater efficiency for the local authorities themselves. But even this argument has a flaw ‘ Wi-Fi mesh in 2.4GHz unlicensed spectrum will not offer the robustness and quality of service to support serious administrative functions, nor to compete with commercial hotspots such as T-Mobile’s network or with 3G data services like Verizon Wireless. The best a metrozone of this type can do is offer a best effort, low cost service with a primarily social function ‘ which brings us back to why cities always deploy first in the downtown areas. Surely these are not just expensive vanity projects, designed to keep the city’s profile high? It is very noticeable that the high sounding talk of social equality and free access that accompanied the original US metrozone proposals has largely disappeared as cities face the need to justify themselves economically ‘ either to give taxpayers good reasons for spending their money on wireless, or to find a profit model such as sharing the system with a commercial provider.
The latter should, while leaving behind early idealism, be a more workable approach. A commercial model will support a higher level of investment over time in creating a robust system usable for more than basic data access by underserved communities. It should also, by providing real competition, spur the telcos into delivering their oft-broken promises to build out affordable broadband, and drive down market pricing overall. As we see major integrators like IBM and Motorola getting serious about metrozones and digital cities, it is clear that a more realistic approach is evolving. IBM itself says that, in the early days in particular, it frequently turned away from bidding on large city projects because there was no clear economic model.
At least companies like Motorola bring real integration and planning skills to the municipalities, where experience of running a telecoms network is low. The giant has signed up with EarthLink, the internet service provider that is aggressively pursuing US metrozone deals, and will provide network design and integration for the five city projects that EarthLink has already won. The partnership is already moving beyond simple Wi-Fi mesh, relying on a combination of mesh networks from Tropos and Motorola’s MotoWi4 broadband wireless system for backhaul and future WiMAX capabilities. The first two projects will be the much vaunted Philadelphia metrozone and another in Anaheim, California.
While we doubt that any model can support yet more Wi-Fi in New York, there would be a case for developing more complex networks that would support differentiated services from hotspots and 3G, not to mention the mushrooming public safety projects. If metrozone projects are to be truly disruptive, they need to be a step ahead of other platforms, not following lamely in their footsteps. And that means leaving behind the assumption that ‘metrozone’ equals ‘Wi-Fi mesh’. Mesh is a fine technology for covering large areas quickly but the use of Wi-Fi radios is increasingly looking technologically stretched and less financially attractive than was once assumed. As the larger vendors enter the market, it is noticeable that the definition of mesh is broadening to include other technologies, even wireline. This is not just to increase the fee the supplier can charge, but because Wi-Fi mesh, before it even reaches its second generation, is becoming stretched at the seams ‘ at least in any market where it faces real competition or where there is dense urban population.
Wi-Fi in 2.4GHz is subject to high levels of interference in any densely populated area. And as the pathetically low usage levels of early hotspots start to rise rapidly, service quality will degrade ‘ driving all but the least affluent (who are unlikely to have Wi-Fi devices anyway) to better supported services. To avoid this, more and more access points will be needed, plus upgrades to support dual-radio meshpoints or VoIP, all of which will damage the economics of the supposedly ‘cheap’ Wi-Fi mesh. Indeed, Motorola, despite its intense interest in mesh, has produced detailed network planning models that show that blanketing a city like London with WiMAX would be cheaper, even at current early stage pricing levels, than doing the same with sufficient Wi-Fi access points to sustain a moderate usage level and quality of service.
WiMAX is one option for improving the mesh proposition, since it can achieve greater range, even in the 5GHz spectrum where propagation of signals is less effective; it lends itself to meshing; and it can support Wi-Fi for end user access so that standards laptops and dual-mode phones can be used. Most mesh specialists have a roadmap to create WiMAX/Wi-Fi hybrid systems from late 2006, and are looking towards the option of a full WiMAX mesh around 2008-9. Advanced hybrid deployments in Finland, Canada and elsewhere, and the sophisticated measurements of companies like Motorola, suggest more forward looking metrozone builders should look to WiMAX from day one.
But municipal networks do not have to be all about wireless, as long as standard user equipment such as Wi-Fi laptops can be supported for access, in order to remove the cost of subscriber units from the business model. Some municipalities are very aware of this and, while using all the wireless buzzwords in their public statements, are actually clearly more interested in municipal fiber.
So the Paris mayor, Bertrand Delano’, hits headlines by promising free access to the internet but, in the fine print, seems keen to base the network on fiber rather than wireless alone. Paris, one of the most competitive broadband cities on earth, may not need another service at all ‘ although French taxpayers would be more tolerant of a publicly subsidized, socially inclusive project than US ones would. But it certainly knows from bitter experience that Wi-Fi is not the best technology to support a metrozone, after the low uptake of its hotspot network based in metro stations and backhauled with the RATP transport authority’s fiber.
All this means that the famous Google approach, as proposed to the San Francisco metrozone project, is in reality less viable than it appears. Google is proposing a genuinely interesting and disruptive model by offering to build the network for free and recoup its costs through internet advertising revenues. The idea appears to provide the city with a risk-free project that will bring free access to many citizens, while enhancing Google’s profile and revenues too. But Google is only promising a 330Kbps service based on 2.4GHz Wi-Fi, and its estimates of how many access points will be needed to blanket downtown San Francisco are hopelessly over optimistic. In other words, under current proposals it will offer a best effort network that will fare poorly against the competitively priced commercial systems available in the hi-tech city. The appeal will be mainly to deprived portions of the community, who are hardly the ideal targets for Google’s advertisers.
The concept is sound then, but the technology is inadequate and the target market ill-thought out. Which is why we believe Google is just putting a stake in the ground with its San Francisco proposition and in reality is looking to invest in, or partner with, more robust networks for the future. It is known to have been acquiring dark fiber in parts of the US and could easily tie up with the owner of a ‘serious’ network such as Clearwire. At this point, the cosy statements about helping deprived communities can be abandoned entirely in pursuit of a fully commercial model that will drive Google’s own business.
Other cities are adopting the advertising supported approach, including Boston, which now offers free access in four commercial districts with costs covered by ads. The city says its Boston Main Streets Wi-Fi project was created with the goal of promoting economic rejuvenation, yet it is focusing on well served areas and, again, has the dilemma of how to address a disadvantaged sector of the population while still appealing to advertisers.
The fact is, metrozones are part of the debate that is as old as government itself ‘ whether services should be invested in by the authorities for the good of the population, or whether they should all be commercially viable. Transport, utilities, now broadband access are part of the same general discussion, and in the end, governments have to make a clear choice, whatever nice words they trot out for election campaigns. The conclusion is likely to be, in most cases, that the networks must be commercially justified for greater city efficiency and/or profit partnerships, and that will require more competitive technology than Wi-Fi mesh. Bringing large private backers like Google into the equation solves many problems with the taxpayers, but inevitably subverts the public spirited aims to those of the commercial provider. In all this, there is a need to think more clearly about metrozones, to decide in which cities they are actually needed before wasting either public or private dollars, and to devise a technology roadmap that goes well beyond unlicensed Wi-Fi.