Original URL: http://www.theregister.co.uk/2007/09/10/municipal_wif-fi_needs_licensed_spectrum/

US municipal Wi-Fi loses its shine

Business models followed technology too closely

By Faultline

Posted in Broadband, 10th September 2007 13:02 GMT

A year ago, hardly a week went by without news of a major municipal wireless project in the US, supporting free or subsidized access and a host of shiny business models. Now hardly a week goes by without news of the death of one of these plans.

The San Francisco roll-out, the poster child for the free Wi-Fi movement, largely thanks to Google’s involvement and genius for publicity, hangs in the balance; now Chicago has cancelled its plans, saying it would be “too costly and too few residents will use it”.

But just as the wilder theories that municipal Wi-Fi would sideline the telcos and kill the cellular business case were patently unrealistic, so talk of the death of municipal networking and non-viability of alternative service models is also exaggerated.

The mistake was to associate the new business approach too closely with the technology. Wi-Fi in unlicensed spectrum was readily available, but was never going to be a sufficiently robust technology to support the ambitions of the Google-inspired internet lobby. This is not to say the concept of an open access wireless internet has been fatally, or even seriously, damaged – it has just shifted to new platforms that stand more chance of delivering an experience that is friendly for the users and profitable for the providers.

Hence Google’s defocusing on muni Wi-Fi and new-found passion for 700MHz licensed spectrum and new devices. The search giant knows, as do Intel, Microsoft and the other drivers in this space, that licensed spectrum and technologies remain the key, at least in the medium term, and that the challenge is to change the rules to support open internet models, new operators and wider availability – rather than try to build an unlicensed new structure from scratch, with the danger of creating anarchy and commercial failure, rather than a brave new world.

There are various reasons why the municipal Wi-Fi phenomenon was always doomed to be an interlude in the history of US telecoms rather than a defining force - which is not to say it has not been important. Like Wi-Fi hotspots, it has challenged the telcos to adapt to a more open world, more rapidly than they would have liked, and it has helped to reshape user expectations of what a wireless internet service can provide. Familiarity with low cost, flat rate, always-available and open access services, even if the experience has been flawed, has certainly hastened the collapse of the walled gardens.

But the early municipal trend was built on idealistic notions of bridging digital divides and opening internet access to all, and was largely reliant on the old-fashioned concept that universal access would benefit society and the economy, and therefore should be funded largely from the public purse.

Such approaches quickly foundered on taxpayer reluctance, and the cities’ desire for blanket networks to raise their own profiles played into the hands of the companies leading the drive for new telecoms models in the US, with Google as their cheerleader. These, of course, wanted to hasten the introduction of services that followed the PC internet approach, providing flat rate, low cost broadband with open access and with the main revenues geared to advertising.

The interest in municipal Wi-Fi mesh clearly offered a readymade opportunity for Google to demonstrate its ad-driven model and find a delivery network for it, one that did not require lengthy and expensive negotiations for spectrum rights.

Amid the enthusiasm for ad-driven wireless businesses, however, it was notable that those with real experience of city networks remained skeptical – the well respected MobilePro, for instance, backed away from free access/advertising projects, claiming these were unworkable. The future of real city networks may well lie with companies like this, building out systems for cities with genuine underserved communities, and for applications with real ROI such as traffic monitoring – in other words, traditional business models resting on a mixture of public funding and provider tariffs.

The problems for implementing the Google approach on unlicensed Wi-Fi boiled down to inadequate technology; high risk spectrum; limited user demand; and conflicting operator requirements. Wi-Fi, for all its merits, is constrained in what it can deliver by its confinement to unlicensed bands, which are subject to interference and poor performance.

Mesh equipment has become more sophisticated in the second generation, supporting voice as well as data and with improved hand-off and routing, but this has also increased the capital investment required. For instance, the general rule in dense urban environments is now to install about 60 access points per square mile, double the number needed for the early data-only meshes. Wi-Fi is certainly not the cheap and cheerful solution it was once portrayed to be, and yet without the security and QoS of licensed bands – or at least licenseexempt bands that are subject to stronger good conduct rules than 2.4GHz typically is in the US – it is difficult to deliver a carrier class experience that would justify high ARPU. ‘Real’ operators remain primarily interested in ARPU, and that will change very slowly.

So the Google/Earthlink model emerged, with the internet players running a best effort, free service supported by advertising, and the experienced provider taking responsibility for managing the network, and offering higher value, tariffed services. On paper this looked fine – and will still work, if a carrier class network in licensed spectrum is introduced for the premium and enterprise services, hence the interest in WiMAX/Wi-Fi combinations among operators like Clearwire. But with Wi-Fi, the carrier was stuck with a network that could not guarantee to deliver premium services effectively. And at the best effort end of the scale, most municipal networks – as Chicago has found – have been launched in city centers where users typically expect better quality, and have access to other free networks.

Here is the fundamental conflict at the heart of the failure of muni Wi-Fi – successful advertising depends on targeting user bases that are attractive to the advertisers, but in many cases these are the least likely bases to take advantage of a low quality service, while the underserved communities that would genuinely welcome free access are ignored, as not being desirable to the clients.

Municipal networks tried to achieve two goals that have proved incompatible in the free market US – though may work better in emerging economies or in some European nations where public funding is not yet a dirty phrase. Chicago’s comments, including a statement that the municipal network would still require “massive public financing” even with an advertising model, show that digital divide projects need tax dollars, and must compete for these with other candidates for the city budget.

The other goal, to introduce a disruptive and open service model to challenge that of the cellcos and broadband giants, will be achieved, just not using city Wi-Fi. The municipal networks have proved a catalyst, heightening interest and confidence in open access and ad-driven approaches, but to attract the broad uptake by a wide range of users that advertisers will need, far more sophisticated offerings will be needed than Wi-Fi could support in 2.4GHz.

Full mobility, location awareness on handsets, advanced multimedia – these will be the bottom line for advertisers in advanced economies, and these need the robustness of carrier class technologies like Wi- MAX and spectrum that has some rules and safeguards. This puts the ball back in spectrum owners’ courts, and some are making innovative moves in their own right, notably T-Mobile USA with its parallel and increasingly tightly integrated hotspot and cellular systems; or Sprint and Clearwire, with their plans to support the open access model, but on their own spectrum.

Google’s mission now is to ensure that bands such as 700MHz have those rules, but rewritten to favour a mobile internet/advertising model that will really work, and generate profit, rather than being just a glorified publicity stunt.

Chicago’s city authority said last week that it was cancelling its plans for a municipal Wi-Fi network because it would be “too costly and too few residents will use it”, and would require “massive public financing”. EarthLink and AT&T were the operators involved in the project, with AT&T dropping plans to build a mesh network for nearby Springfield, Illinois.

EarthLink has been pulling back rapidly from its once fervent commitment to city Wi-Fi as a means to build itself a broadband base that would bypass the telco networks. Last week it said it was cutting 900 jobs from a total workforce of about 2,000 and closing several offices. It has eliminated the post of president of municipal networks and says it will now only commit to municipal projects if the city pays for the construction. "Where we are building or have built networks, we will continue under the old business model," said a spokesperson. "But in cities where there are no signed contracts, we are revisiting the issues." Officials in 12 cities in negotiations with EarthLink were not available for comment, but the ISP has already pulled out of the much vaunted San Francisco plan.

This could spell the end of Google’s aim of making San Francisco the showcase for its free access model – it was EarthLink’s primary, but increasingly silent, partner in the project. The company did reveal some results from its own free muni Wi-Fi system, in its native town of Mountain View, California, and reported that the network now covers 12 square miles, based on 400 mesh routers, and counts 15,000 unique users each month. It said traffic has grown by 10 per cent per month and the network now carries 300Gb of data every day.

The slowdown in municipal build-outs will have a knock-on effect on the mesh vendors, and growth is slowing, although it will still be about 90 per cent in 2007, according to In-State. The researchers say that the Wi-Fi mesh equipment market had more than 100 per cent shipment growth in 2006, and will have more than 90 per cent growth this year, and that strong growth will continue for mesh access points for the next several years, as shipments grow more than threefold between 2006 and 2011. However, most of that growth will come between 2006 and 2008, with rates rapidly declining in 2009. “Cities will continue to deploy municipal mesh networks, but the rate of new deployments after 2008 will slow, due to concerns over the business model,” said analyst Daryl Schoolar. This means mesh vendors will start to focus more on enterprise systems – a sector some companies like Strix sidelined to chase the municipal boom – and on introducing new technologies like WiMAX.

Copyright © 2007, Faultline

Faultline is published by Rethink Research, a London-based publishing and consulting firm. This weekly newsletter is an assessment of the impact of the week's events in the world of digital media. Faultline is where media meets technology. Subscription details here.