The fervor surrounding Wi-Fi metrozones is reaching fever pitch, with almost a city a week announcing plans in the US, and key European locations like London following suit. But the supporters should cast their minds back just a few short years, to all the sound and fury that surrounded the early roll-out of Wi-Fi hotspots. A similar pattern will appear in metrozones ‘ over ambitious hopes; a failure to find any profit model despite the undoubted benefits of the services to end users; the telcos and cellcos asserting control. The alternative operators will need to learn the lesson of the hotspots ‘ that there is no money in Wi-Fi unless it drives other services and revenues, and that large players will dominate in the end. Like T-Mobile in that first wave, the Earthlink/Google/Motorola combination stands out as the body that has recognized that fact, and could therefore provide a real alternative to telco control in many large cities.
Hotspots were undoubtedly a great phenomenon for laptop carrying users looking for (relatively) cheap, convenient access; also a great bonus for makers of those laptops and their chip suppliers. Less exciting for the operators of the hotspot services. Amid talk that a global network of hotspots would eliminate the need for 3G mobile data services, the grandest of the start-up initiatives that bought into that dream quickly shrivelled (remember the IBM/AT&T venture Cometa?) The main gainers were the aggregators, notably iPass, but while hotspots drove some additional revenue and reduced churn for some venue owners and WISPs, they failed to find a profit model in their own right, as they came up against the basic user expectation, engendered by the internet revolution ‘ that they should have ever faster and more ubiquitous broadband access, for ever lower, and certainly flat rate, prices. With their entire business model under threat from this trend ‘ not specifically from hotspots or any single technology ‘ the cellular providers bit the bullet and took control of the hotspot trend, with T-Mobile, Swisscom and others blazing the trail. Overall data margins were squeezed painfully and with wireless VoIP, the voice cashcow looked to go the same way, but at least what money there was in public Wi-Fi was safely back in the cellcos’ pockets.
The same pattern is clear in metrozones, but on a grander scale, because Wi-Fi (or in future, WiMAX) mesh can provide coverage similar to that of cellular networks and second generation meshes are supporting VoIP and video as well as data. This means the threat to telcos is greater, and after an initial attempt to resist the onset of the metrozone, this has prompted many of them to try to wrest control of the trend. IBM, heavily involved in metrozones on the integration front, has seen a noticeable move in the US away from idealistic city-backed projects aimed purely at public safety and bridging the digital divide, towards systems that will generate profits and carry premium services as well as best effort access, and which will be driven in many cases by existing operators. Some of these will be, like EarthLink and cable operators like Cox, wireline providers looking to reduce churn in their base and create a wireless base without a 3G license. By partnering with companies like Google, which will fund a best effort service in return for increased advertising revenue, these operators can focus on delivering premium services for a profit, and biting into the telcos’ market share.
Many large US telcos are now targeting a piece of the metrozone pie. AT&T put in a bid this month to provide the wireless internet service for Michigan’s Washtenaw County with roughly 325,000 residents. Cableco Cox Communications recently teamed up with two companies to offer wireless access in some Arizona cities (see Wireless Watch March 20 edition), and Time Warner Cable has signaled interest in Texas. AT&T is working with Tropos and IBM in Michigan and is proposing offering at least five hours of free service per month at DSL-like speeds, and unlimited free access at slower speeds. This shows the real dilemma for the large telcos ‘ by joining in the metrozone race, they are undercutting their core broadband access prices, but without the compensation of a revenue stream such as advertising, as Google has; but as the limited success of anti-municipal legal actions showed last year, if they remain outside the game, they will lose revenue and margin anyway, but to competitors. AT&T is also bidding in Michigan’s Genesee County.
Time Warner is bidding in Dublin, Ohio and Corpus Christi, Texas ‘ also of interest to AT&T ‘ while Comcast’s venture arm has invested in mesh specialist BelAir Networks, which has created a product that integrates Wi-Fi access points with cable lines, so that cablecos do not have to pay for access to infrastructure such as light poles.
The latest telco to get involved in this race is Sprint Nextel’s soon-to-be spun off local phone arm, now renamed Embarq. Cut free from its parent, Embarq will face all the threats to its core voice revenue that the larger local operators have, and like them it is turning to wireless to compensate, with plans for a cellular MVNO (presumably leasing capacity from its former parent) and to bid for metrozone contracts. It has set up two square-mile trial networks in the Las Vegas, Nevada suburb of Henderson, one of the fastest growing cities in the US and one of Embarq’s largest markets. Using BelAir equipment, Embarq aims to create a business model that would reduce churn and enhance the service bundles it can offer its traditional voice base.
Such initiatives by Embarq could be included into the complex wireless ecosystem that Sprint Nextel and its cableco partners are creating, combining its 3G and forthcoming broadband wireless networks with cable content and fixed customer bases, and a new interest in metrozones. Sprint Nextel recently bid for a project in Grand Rapids, Michigan, its first major sign of interest in adding municipal access to the patchwork of technologies with which it will seek to surround and undermine the Bellco incumbents.
Of course, once Sprint has built out its broadband wireless network in its plentiful 2.5GHz spectrum, it will probably find Wi-Fi meshes redundant in most of its key cities, since the new network will have sufficient bandwidth to support various classes of service, with the quality of licensed spectrum. Other telcos are thinking in these terms too ‘ knowing they have the spectrum and the funds to invest in more advanced technologies than license-exempt Wi-Fi, and so the potential to trump the metrozones created by smaller upstarts and even EarthLink (though the latter has the weight to negotiate its own spectrum deals, potentially with Clearwire). BellSouth, which has rolled out Navini pre-WiMAX technology in a few cities, said last week that it would test WiMAX for larger citywide deployments.
BellSouth, soon likely to be part of AT&T, is finding various ways to generate revenue from its 2.3GHz spectrum and drive additional revenue from its other networks. For instance, it plans to use its broadband wireless to offer a back-up solution for businesses and consumers whose broadband connections fail ‘ a plan that builds on the emergency back-up services BellSouth provided, using Navini Ripwave kit, after the New Orleans hurricane, but is now being applied to regular back-up. The service will cost an additional $30 per month in addition to a broadband subscription. BellSouth has been offering wireless broadband in the 2.3GHz band as a standalone service for $70 per month in several southern cities. ‘Katrina was a real life lab for us when it came to developing back-up strategies,’ said Michael Bowling, vice president of convergence and platform development at BellSouth. ‘But we found that customers wanted to return to their regular broadband service when things were up and running again.’
Even Verizon’s chairman Ivan Seidenberg, who has been outspoken against WiMAX and Wi-Fi, is buying into the notion of future broadband services based on a combination of different technologies. Speaking at the TelecomNEXT conference in Las Vegas, he said: ‘The new internet revolution will be characterized by 100Mbps two-way capabilities, and broadband-anywhere mobility. It will be delivered by an increasingly diverse range of technologies – fiber, EV-DO, cable, Wi-Fi, WiMAX – from a wide range of providers.’
EarthLink and other large service providers are banking on the unwillingness of city authorities either to own and run wireless networks, or to spend tax dollars on them. For instance, the government of Minneapolis, Minnesota recently decided to abandon an original plan to own its own metrozone ‘ a plan driven by the desire to ensure access to underserved communities – and instead contract with an ISP to run the system as a commercial enterprise, with no safeguards announced for universal access, such as subsidies for poorer citizens. Social justice might have dropped down the agenda, but it will be good news for one of two providers in the running, the rampaging EarthLink (whose equipment/integrator partner is Motorola, supported by mesh specialist Tropos); or the locally based ISP US Internet. EarthLink’s latest deal is to create a metrozone for Milpitas, California, covering 6.5 square miles. This will support the now common EarthLink model of being open to multiple service providers, with the company itself focusing on high end and business offerings and other ISPs offering consumer access.
The US is starting, with initiatives like EarthLink’s, to work out some of the early challenges of the municipal network, particularly learning the lesson that there is likely to be no real money in these networks per se, unless they can drive additional revenue for operators in other areas ‘ and that means attracting the larger service providers, or the internet players like Google, even at the cost of early dreams of social justice. In Europe, the thinking is lagging at least a year behind and some of the mistakes are set to be repeated, especially given that the need for metrozones in western Europe is arguably less than in the US, because of the greater penetration of cellular technologies and the mobile culture. Undeterred, projects are going ahead even in the most mobile cities ‘ including Paris and several overlapping initiatives in London. The UK is now second only to the US in terms of urban wireless build-out, with public sector-backed metrozones due to go live in the next few months not only in the capital but in the cities of Birmingham, Cambridge, Edinburgh, Leeds, Liverpool, Nottingham and Oxford, among others. However, in almost none of these cases is there a clear business plan, and we can expect to see the systems either being mothballed or declining in quality as resources are diverted elsewhere, or being taken over by the telcos, with companies like BT keen to compensate for its lack of a 3G license by getting involved in other wireless activities such as hotspots and WiMAX.
It is not BT but a start-up called Urban WiMAX that is bringing 802.16 systems to the UK capital first, however. The company has announced a fixed wireless service aimed at small and medium enterprises in the Westminster section of central London, emulating other city-based broadband wireless operators such as Towerstream and Libera in targeting leased line or DSL replacement for businesses as the initial, relatively safe business model, before expanding into consumer access.
The £4.4m deployment will deliver broadband speeds of up to 10Mbps, both upload and download, to over 250 businesses that have signed up for a free wireless connection in the trial period, beginning next month. Full commercial launch of the service is expected in the third quarter. The London-based company recently agreed a three-round funding arrangement for £16.95m, £1.7m of which is earmarked for the first phase of network roll-out. The company says the service is typically five times faster than business DSL, which often has a 2Mbps ceiling at the same price point, and 50-70% cheaper than leased T1/E1 lines. The aim is to find ‘a way out of the DSL straitjacket’ for small businesses, the company argues, and will support IP Centrex VoIP by using a dedicated pipe with an encrypted channel over WiMAX.
Urban WiMAX’s new service will be rolled out in other London areas where there is a concentration of SMEs in suburban districts, and the company also plans to roll out fixed WiMAX in 12 major UK cities over the next 12 to 18 months using 5GHz unlicensed spectrum and possible new options coming from regulator Ofcom in the future.
Its business plan, while it will face stiff competition, is at least cautious and based on well established market demands. The metrozone initiatives expanding in London and elsewhere show no such good sense, and there will be considerable pain for operators in these areas, even while consumers and smaller businesses reap the benefits of low cost access. Like the hotspots, in the end this will be a game for big operators, whether traditional or coming in from other sectors.
Latest metrozones
Rio Ranco, New Mexico will soon offer citywide Wi-Fi internet access to every resident and visitor for free, for a best effort 100Kbps service limited to 10 hours per month. After that, additional hours or faster service, up to 1.5Mbps, will be charged for. Its network provider is Azulstar Networks of Grand Haven, Michigan, using Meru kit. The city has decided to forego any share of the revenues in order to encourage uptake and ISPs. It has been backed heavily by Intel, which has plants in the area.
Ricochet, a failed wireless provider of the last boom of 2000, is trying the latest in a series of comeback bids by moving into metrozones. Its proprietary mesh-like equipment remains installed in many parts of Silicon Valley, though often disused, and it is bidding to use this equipment to create metrozones in San Jose, among other cities.
SkyPilot Networks and MetroFi have announced that the City of Aurora, the second largest in Illinois, has selected them to build out the first citywide Wi-Fi network in the state. The mixed-use network will provide free public access as well as new public works and public safety services. The partners have already won contracts in Cupertino, Santa Clara and Sunnyvale in Silicon Valley.
Sacramento, the state capital of California, is working with MobilePro to create a citywide network using up to 1,000 access points from Strix Systems. Mobile Pro will own and maintain the network, and will act as a wholesaler for other voice, video, and data providers that wish to run their services over the network. The target is 10% penetration of the city’s broadband market, with other providers including Covad, AT&T, the WiMAX start-up Speakeasy and Comcast. The city will provide infrastructure and, in return, will receive free wireless broadband service for use by city employees. Strix’s gear can be upgraded to support WiMAX by adding a network card, said MobilePro.
Tropos upgrades
Despite its success in US metrozones, especially through partnerships with Motorola and IBM, some question the medium term viability of Tropos’ mesh hardware because it uses a single Wi-Fi radio, while most rivals use a second radio for backhaul or dual-band support. However, the vendor claims its approach keeps costs low and it has now boosted capacity in its MetroMesh system too, with a software upgrade that promises an average of 20Mbps per square mile, in terms of total bandwidth available to all users on the network at the same time in that area, up from 15Mbps.
The hike comes from automated channel allocation of MetroMesh nodes. Channels are picked based on information gathered from all the nodes according to interference levels and dynamic channel switching to avoid ‘noise’ is now included in the vendor’s Predictive Wireless Routing Protocol (PWRP).
Also new in version 5.1 of this software is support for collecting data for new tools -Tropos Insight for remote data analysis and multimedia optimization and Tropos Drive, a hardware appliance used by network administrators as they physically drive around a Tropos mesh, gathering data on the throughput of the network.
MetroMesh 5.1 will also help ISPs integrate wireless mobility ‘ important to keep Wi-Fi competitive against WiMAX and cellular ‘ by supporting GRE tunneling and BGP routing, used by many network operators for user authentication. Logged on this way, users can roam from network to network without losing a signal.
Mesh growth trends:
More than 1m wireless mesh routers, generating revenues of more than $1.2bn, will be shipped in 2010 to service municipal Wi-Fi networks, market research group ABI Research concludes. By 2010, municipal WiFi networks will cover 126,000 square miles worldwide, an increase from about 1,500 square miles in 2005.
Meanwhile, Juniper Research estimates that global spending by local government bodies on mobile and wireless solutions last year hit $802m and is set to reach $8.6bn by 2010.
Metrozone roaming:
As in the hotspot world, one way to enable smaller providers to remain in control of some metrozones will be roaming and shared services. The aggregators are starting to include metrozones in their catalogs ‘ some, such as iPass, focus on providing unified access, security, sign-on and billing to enterprises, regardless of which participating hotspot network their employees use; others, such as Roampoint, also negotiate roaming deals between the various network owners and act as a hub handling inter-provider billing, authentication and so on. Membership of one of these super-networks enables smaller providers to survive and attract additional traffic. A recent addition has been SparkNet, which has moved into the metrozone market by offering back office and roaming services as well as building networks for city or private operators. Its OpenSpark community of networks is mainly focused on Europe though it has some US participation, and most recently added the city of Turku in Finland, creating a metrozone built around 500 access points.
Clearwire has raised further funding, with $360.4m in a debt offering, and is likely to have topped the $1bn mark by now. It also looks increasingly likely to create a mirror strategy to that of Sprint Nextel, the other significant holder of US 2.5GHz spectrum, but partnering with satellite providers rather than the cablecos favored by the larger company.