Mobile operators have had a decade to deal with the knowledge that they could no longer control or monetize most of the services running over their networks, as the world moved to over-the-top offerings. Now they have to cope with the additional realization that they may not control the networks themselves forever either. Of course, the dream of a market with no cellcos, no walled gardens and no long term spectrum licences will remain a dream for many years, perhaps indefinitely. But the share of wireless services which ride on cellco-controlled networks will fall, putting pressure on access and value added service revenues.
Mobile operators have generally played a strong hand in bringing the anarchic public WiFi under their own control via ‘carrier WiFi’, offload and Hotspot 2.0, but that was only the first skirmish in the war to control the networks. Now they face attacks on three key fronts ‘ telcos and cablecos harnessing WiFi as an alternative to cellular deals; Google building its own infrastructure and threatening to create a huge WiFi cloud of its own; and various firms building dedicated networks for the internet of things (IoT), one of the cellcos’ biggest hopes for new LTE revenues.
Events this week have taken place in all three of these areas of threat to cellcos’ ROI plans for LTE.
Google to create US-wide business WiFi network
A decade ago, Google was vocal in the thinking that Wi-Fi hotspots, meshed into dense urban networks, could be a viable alternative to cellular (for data at least ‘ even now, mobile voice really is a step too far). The business case for the much-vaunted ‘muni WiFi’ didn’t really work out, and despite its own build-outs in Silicon Valley, Google did not emerge as a mass-scale hotspot deployer and slayer of cellcos.
But that doesn’t mean it left its old visions behind. The search giant is investing heavily in building networks, mainly in the US and some emerging markets, using fiber, balloons and unlicensed wireless. It does not want to be a traditional network operator, but to showcase a new model which could, it believes, will bring ubiquitous broadband access to the world’s population more quickly, and with lower usage fees, than the old cellular approach.
Because it controls the platforms, it can use them as it wishes, showing off ideas such as supporting hundreds of small service providers on an on-demand basis, rather than with static MVNO deals. This is not about broadband revenues for Google, but about encouraging a wireless model that will make its web services available and affordable to everybody, expand its ecosystem, and so driven its core business. Cellular networks will still be important, but increasingly, these new-style systems will be a complement and a spur for changes in regulatory and carrier behavior.
After the muni WiFi bubble burst, WiFi seemed to play a smaller role in Google’s networking initiatives, though it continued to be a powerful lobbyist for new WiFi spectrum such as the TV white spaces. But the WiFi technology has moved on, and is now far better equipped to be part of the alternative network. It has got faster, with the 11ac and 11ad extensions; and it supports seamless integration with cellular networks via Hotspot 2.0, Passpoint and NGH (Next Generation Hotspot). The latter developments help cellcos harness WiFi for themselves, of course, but it also makes it more practical for WiFi-only, or WiFi-first, models to thrive (since any service claiming ubiquity must, at least sometimes, hand off to cellular).
And in many areas, despite the importance of WiFi offload to mobile carriers, it is the non-cellcos which are deploying new generation WiFi to greatest effect. One of the turning points has been the emergence of the homespot, as pioneered by FON. This is a WiFi home router which has two SSIDs, one left open for passers-by to use for broadband access (consumers who agree to this get an incentive such as lower rates or free content). The homespots are a very low cost way for broadband providers to create clouds of coverage and capacity, especially in residential areas where the hotspot model might not make sense. They are being used disruptively by some of the most threatening (to the cellco) providers, such as Comcast in the US ‘ leader of the huge CableWiFi hotspot alliance ‘ Free in France, and many wireline providers like BT in the UK. Some of the homespot enthusiasts, like Deutsche Telekom, Telstra and Free itself, have mobile arms too, but recognize that they can bolster their fixed broadband propositions, and move more cheaply to quad play, if they bet heavily on WiFi.
Building a huge network by offering cheap gear to residents and cheap access to travellers has been proven in the consumer space, as a low cost way to achieve wide area coverage. Google is now reportedly putting a twist on that concept and addressing businesses, offering subsidized WiFi routers in order to create a US network of hundreds of thousands of nodes, which it will manage from the cloud. They will also get low cost or free access provided they join Google’s public network ‘ another example of how the search giant will make its money back via a bigger network for its adverts and applications, not via access fees as in the traditional operator model.
According to GigaOM, Google will use Ruckus’s newly announced cloud-based WLAN controller system (see separate item) to help scale up its business WiFi network quickly. The idea is to build a platform to which any business nationwide (and perhaps beyond the US in future) could connect its WiFi routers, gaining easy deployment and management, and low cost equipment. In effect this would add the small and medium businesses, which are the main target, to the clouds of residential WiFi already being created via the homespot roll-outs, and to public hotspot systems like CableWiFi.
Ruckus’s virtualized system can connect tens of thousands of access points to the same virtual network, which will save Google the complexity of managing many local business WLANs individually. It can also use the platform to deliver cloud-based applications such as advertising and point-of-sale payments in a centralized way, to all the connected access points; and it can sell services such as analytics to the participants. Sources said Google was only working with Ruckus, though other WiFi companies such as Cisco, Ericsson/BelAir, Aerohive and Cloud4Wi also provide cloud-based management (though none of these systems is interoperable). The service will rely heavily on Hotspot 2.0 so that users can log in automatically to any Google access point and can hand off to cellular networks if WiFi is unavailable.
This is just one way that Google hopes to push WiFi to its limits in terms of supporting its core businesses. It is already working in a more traditional way in WiFi service provision, and has replaced AT&T as the WISP for Starbucks coffee shops in the US. It is building free WiFi zones in US public parks and in white spaces spectrum in Africa. It is also active in evolving the technology itself, so that it is able to take advantage of any new unlicensed spectrum which comes available, such as 60GHz ‘ the firm has an experimental licence to test emerging gigabit WiFi and HetNet/small cell technologies on its Mountain View campus, for instance.
Google may have played nice with the cellcos for the past few years, in order to get Android out on the cellular networks and ensure a strong platform for its mobile services. But it clearly still believes it could make even faster progress with a new, open approach to wireless broadband, one which will be driven by web giants like itself, Facebook and Amazon. This is not about killing the cellco, but rather about supplementing their networks, and eventually making them irrelevant.
Homespots drive disruption from US to Australia
While Google expands on the FON model in the business sector, the original open WiFi proponent continues to sign new partners, most recently Australia’s incumbent Telstra. Always in the vanguard of new wireless technologies, Telstra has announced plans to build a national Wi-Fi network, at a cost of over AU$100m. While it will build about 8,000 urban hotspots between 2015 and 2018, it will boost that figure to over two million access points via FON’s homespots.
This will create a massive community of WiFi locations offering 2Mbps download speeds on the public side, while having no impact on the user’s own broadband speeds. Fixed line Telstra subscribers will have free access to the network though downloads will count towards the quotas included in their home line contracts.
Telstra has experimented with public WiFi before, but closed its network in 2007 because of low uptake. But, like other operators, it has seen demand and technology move on, leading to a more viable redefinition of ‘muni WiFi’. The company said its original network used 802.11b, but now it will use 802.11ac for its own hotspots. It will spend AU$50m on those plus core infrastructure, and a further AU$50m on partnering with other hotspot owners.
CEO David Thodey insisted there was no conflict with the cellular network (though WiFi does reduce overload on 3G/4G, and therefore the amount of capex required to help it keep up with demand). “The cellular network is great if you’re on the move, but if you’re sitting down watching a few movies, it’s a different type of access,” he said. “We still think the cellular mobile network will have a great future. We just want people to have a great connected experience.”
FON’s COO Alex Puregger echoed these views, saying: ‘Working with Telstra is a huge achievement for us. Not only because it means that we are now expanding our network to a new continent, but also because Telstra has been a global leader in LTE deployment. It emphasizes our view, that LTE and WiFi are complementary technologies, both needed to address future growth of mobile data demand.’
Meanwhile, cablecos round the world, and particularly in the US, have no such delicate balances to draw between WiFi and cellular. WiFi hotspots and homespots are the route to lower MVNO fees to cellcos, added value for broadband customers, and a national wireless data network under their control. Cablevision, the first major US MSO to embrace this idea, said its aims was to ‘disrupt’ the mobile data market and CEO James Dolan, on the firm’s recent earnings call, hinted at some big new launches ahead, though he offered no details. “We’ve been saying for quite some time that WiFi is a differentiator for our business,” he said. The company’s Optimum WiFi network has more than 100,000 access points and he promised: “I think you’re going to see new products, some of which will be disruptive.” Like Comcast, Cablevision is part of the CableWiFi national roaming deal, and is installing homespots to expand its network affordably in its New York area territory. With homespots, it aims to reach one million access points by the end of this year, while Comcast is targeting eight million (and also hopes to acquired Time Warner Cable, which had 33,000 at last count).
WiFi is particularly important for cablecos as most are increasing broadband subscribers and market share, while losing video users. For instance, in the first quarter of this year, Cablevision added 8,000 high speed data customers while losing 14,000 video subscribers, becoming the third major US MSO to have more broadband subs than video (Time Warner and Charter have already crossed that Rubicon).
Such trends will drive cablecos to invest even more heavily in WiFi in order to lure data users from the telcos and even the LTE networks. Comcast is even looking for global rights for its subscribers, signing trial roaming deals with KDDI in Japan and Taiwan Mobile. It would be logical for the MSOs to look to improve their video performance by using WiFi (and, secondarily, cellular MVNO deals) to offer an integrated quad play bundle. Many are speculating that Comcast, in particular, will make this move and offer a ‘WiFi-first’ smartphone (which defaults to WiFi and only moves to cellular if none is available).
At the recent NCTA conference, however, cable executives remained cagey about such plans. On a panel about Cable WiFi, Leo Cloutier, SVP of corporate strategy and business development at BrightHouse Networks, did say this: ‘We’re all very aware that possibilities exist as we build out more WiFi. What you’re hearing from us is, none of that is a reality now. None of that’s something we’re doing now. We’re aware those possibilities exist. We’re building a useful WiFi network.’
There are many potential uses for the CableWiFi network. MSOs can use it to offer a mobile service to subscribers that could be cheaper than those from cellcos; create a ‘reverse MVNO’ revenue stream, in which the cablecos sell access to the WiFi networks to cellcos looking to offload; offer their own WiFi-first phones; partner with a third party WiFi-first start-up, such as Republic Wireless, to take on the cellcos and challenge any potential telco quad play offerings
But others were even less forthcoming. Cablevision’s SVP for video product management, Brad Feldman, said: “We are using this to help us differentiate from the telcos. We are investing in WiFi and we’ll see where it goes. But we are not a direct competitor.”
And Tom Nagel, general manager of wireless services at Comcast, added: “Wi-Fi is not a licensed technology and it will never be. It’s about nomadic sessions [not full mobility]. They are very different.”
Are the cablecos being disingenuous with such statements? The scale of their activities certainly suggests more than just a useful value-add for customers, and with Google upping the ante itself, they may have to redouble their efforts in order to ensure that they, not the web giants, are the ones to disrupt the cellcos and create a new wireless model for the US.
Sigfox brings its dedicated IoT network to UK and US
Another area where cellular operators will have to adapt their traditional models radically to stay in the catseat is the emerging internet of things (IoT). They may dream of running the IoT entirely on their LTE networks, generating new revenues from their 4G spectrum, but the real picture will be far more fragmented. Initiatives like Japan’s WAUN (wide area ubiquitous network), and companies like Sigfox, Senaptics and Neul, show the potential for dedicated IoT networks to be an important part of the jigsaw.
The M2M market from which IoT springs has always relied heavily on dedicated networks, providing security and specific functionality for individual verticals, and often run by the industrial groups themselves. That thinking is behind growing interest in building specialist, optimized networks to carry IoT services, using ‘ultra-narrowband’ (UNB) technology.
Sigfox has used this technology, in unlicensed bands, to create a dedicated IoT network in France, and is now partnering with UK telco infrastructure company Arqiva to extend its model across the Channel. Arqiva says this will provide ‘nationwide low power connectivity for the first time’. This comes just weeks after another UK venture, Senaptic, announced plans to use UNB to build specialized, ‘cellco-free’ IoT networks for specific customers and segments.
Arqiva has been extending its core business in operating cell and broadcast towers, with activities in growth areas such as small cell infrastructure and Wi-Fi hotspots. This sees it building a broad infrastructure platform covering all kinds of communications sites, and the deal with Sigfox could add vertical and industrial service providers to its mix.
Arquiva will build and run a national IoT network, initially rolling out the UNB technology in 10 cities – London, Birmingham, Bristol, Edinburgh, Glasgow, Leeds, Leicester, Liverpool, Manchester and Sheffield. It will use technology licensed from Sigfox and its system will be connected to the French firm’s own networks. These are currently deployed in France, the Netherlands and Spain, and in individual cities elsewhere, such as Moscow and Munich. Sigfox took a ‘build and they will come’ approach in its home country, but elsewhere is mainly working through partners like Arqiva, as well as creating a global distribution channel for complaint devices, to encourage further uptake.
As at Senaptic, Arqiva’s key target application is the smart city, although the platform could be suited to many segments and enterprises, it said. UNB can run in any unlicensed ISM spectrum such as 2.4GHz, 433MHz and 900MHz, though operators are tending to use the lower frequencies rather than the congested WiFi bands. They cannot claim the high security of a licensed, private IoT spectrum, but argue that their networks are designed specifically for this market, unlike LTE or WiFi, and can run in underused frequencies to minimize interference risk.
Arqiva said in its statement: ‘Sigfox’s ultra-narrowband technology is particularly suited to connecting objects over long distances where a long battery life and low cost are required. By becoming the Sigfox Network Operator in the UK, Arqiva has strengthened its position as a leading provider of machine-to-machine connectivity services.’
Wendy McMillan, managing director of smart metering and M2M solutions at the company, said: ‘We already run satellite, WiFi, and long range radio networks, providing full managed services where required. Our smart metering communications service, connecting 10m homes using long range radio technology, will be one of the UK’s largest machine-to-machine deployments.’
Rodolphe Baronnet-Frug’s, VP of network and business development at Sigfox, said in a statement: ‘According to some forecasts, there will be 50bn connected devices worldwide by 2020, but for this to become a reality, both cost and energy use will have to come down.’
Sigfox is also working directly on its first US project, which will create what is being billed as the first dedicated IoT network in the country (though there are plenty of specialized, private M2M systems). The new build-out will span the hi-tech areas of San Francisco Bay and Silicon Valley, using the UNB technology and the ISM spectrum band.
Sigfox has grand plans in the US, saying it will bring its technology, directly or with partners, to the top 10 US markets over the next 12-18 months. Other cities may include New York, Los Angeles and Seattle. The networks will be open to all kinds of IoT service providers, with bandwidth charges no more than $1 a month per device, or even $1 a year for some very small or low-usage sensors. For consumer devices, this will often be included in the price of the gadget ‘ a model that hasn’t taken off in consumer devices like tablets (except Amazon’s Kindle ereaders) because of their heavy data usage, but which makes sense for low-data products such as smart thermostats and meters. One of the ways to encourage uptake of such gadgets will be to ensure they do not come with additional data fees and plans.
Other types of devices will be sold at low cost to drive a subscription model. In this area, Sigfox’s first customer in California is Whistle, which makes an activity monitor for pets. This will support GPS and UNB rather than 3G or 4G ‘ which would have made the gadget 3-4 times larger, according to CEO Ben Jacobs. A cellular version would have had to incur higher service charges (the GPS rate will be $5 a month) and it would have had far shorter battery life.
Another UK company, Neul, is working on creating a non-cellco, ultra-low power wireless network for IoT applications. It has worked on trials with BT and could support specialist services for mobile operators or directly for enterprises. It has its own technology, Weightless, another indicator of the push for networks that are developed specifically for the requirements ofms of low energy gadgets. The challenge for the cellcos now is to try to ensure they have a powerful role in the value chain, even in markets where LTE does not prove to be a dominant solution.
This week, BT and Neul joined a consortium of companies building a citywide M2M network in the city of Milton Keynes, which will also serve as a demonstration system for Weightless and a testbed for firms developing new applications and services for the M2M platform.
Alan Ward, head of corporate ICT practice at BT, said in a statement: “We see this exciting project as a means of establishing an open innovation environment to support the creation of M2M and IoT applications across a whole city. This could include anything from intelligent monitoring of parking spaces in the city to networked bins which signal when they need collecting. The project will showcase BT’s capability in managing network services for these applications which will be pivotal in making this initiative a success.’
Geoff Snelson, strategy director at Milton Keynes Council, said: ‘As well as providing a testbed for our own specific use cases, this will bring new innovation and business development to the city, creating an ecosystem of IoT development.’