« July 2007 | Main | October 2007 »
August 01, 2007
The Mobile Operators' Wi-Fi Imperative
In just eight short years, Wi-Fi has evolved from home networks, to corporate LANs, to coffee houses and book stores, to corporate and university campuses, and more recently, to trying its awkward hand at metropolitan areas. As it crossed into each of these new domains, it had to scratch and claw to gain respect. It had to constantly innovate to prove that it was up to each new challenge. In the corporate/enterprise market for example, it took several years for IT managers to be convinced that Wi-Fi could be deployed in way that was secure and manageable from an enterprise standpoint. The whole family of 802.11 standards had to evolve as it faced each new challenge.
In its newest domain - metropolitan area networks - skeptics point out the many ways that Wi-Fi as a technology is flawed and inferior to other technologies, including WiMAX and various 3G standards. Of course they are right. Wi-Fi is hamstrung by its reliance on unlicensed spectrum; its lack of quality of service mechanisms; its low power restrictions. But what these skeptics fail to realize is that there’s more to determining whether a technology succeeds or fails than its level of technical superiority. The ditches of the tech-industry are littered with technologies that were better than their competitors, but failed to get into the “disruptive hall of fame” of mass-market, ubiquitous, de facto standards. In almost all cases, these “superior” technologies were not outright destroyed, just relegated to niche markets and applications. Let’s look at a couple of examples.
In the 1990s, ATM was predicted to be the successor to Ethernet. ATM supported quality of service, ultra-high bandwidth and other capabilities that made it ideal for carrying the new latency-sensitive, converged voice, data and video traffic of the future. There was no way a best-effort technology like Ethernet could survive against this elegant new ATM technology! Ethernet was designed as a LAN technology, right? Well, Ethernet didn’t give up. What was once a technology for connecting PCs in small offices at 3 Mbps evolved to 10, then 100, then 1,000, then 10,000 Mbps. And Ethernet is now used to power multi-gigabit metropolitan-area networks. And work on 100 Gbps Ethernet is underway.
Sound familiar? Of course it does; it’s the same debate going on now in some circles about the relationship between muni Wi-Fi, WiMAX and 3G standards. WiMAX uses licensed spectrum and has quality of service. It was designed from the ground up as a metro-area technology. Wi-Fi was built for LANs.
Another famous example; In 1981, the PC was inferior in all ways to the minicomputers it would ultimately replace. Who would have predicted at that time that PCs would become the dominant computing platform in the world? At least one company did; Microsoft.
David Isenberg, in his now-famous paper Rise of the Stupid Network , got it right in 1997. David observed that a paradigm shift would occur as devices - the end-points on the network - became more intelligent. And network-centric quality of service limitations could be mitigated by these intelligent devices, and also by huge increases in the raw bandwidth (capacity) of the networks. So as not to butcher David’s great paper with my paraphrasing, I’m better off quoting a passage.
It is rare that a market is completely killed by the next generation of technology. Neither TV nor the VCR killed the movies. Neither the minicomputer (alas, remember them?) nor the PC killed the Mainframe. We still have ships and railroads, though their markets are both diminished and changed by the car and airplane.
Kevin Werbach, Associate Professor at Wharton Business School, nailed the paradigm shift that occurs when unregulated spectrum meets intelligent devices in his 2004 paper , where he wrote:
Think about a group at a cocktail party. Many people can hold conversations with one another simultaneously. They can do so not because they each shout over the others, or because they agree on a set of rules to define who can talk when and how. It’s obvious to us that the reason so many people can talk at once is that the speakers modulate their volume and the listeners use their brains to distinguish their partner from the ambient noise.
So, the argument that Wi-Fi (or any wireless technology) is “hamstrung” by its reliance on unlicensed spectrum assumes something critical that is no longer true; that the end-point devices communicating over this shared spectrum are dumb listening devices; like an FM radio. This is a dangerous assumption to hold onto if you are in the wireless voice and data business. The 802.11n standard is precisely focused on using better radio and antenna technology to distinguish between intended signals and ambient noise.
Finally, it’s worth noting that Clayton Christensen, professor at Harvard Business School, and the father of disruptive innovation theory, also picked up on the potential for Wi-Fi to disrupt mobile operators. In this article, he notes:
Ironically, the wireless industry’s history traces back to disruptive innovation. The first mobile phones had low voice quality, limited battery life, were bulky and expensive. But they offered something that land line telephones could not match: The capability of placing and receiving calls while mobile.On its own, the growth of Wi-Fi threatens wireless carriers who are increasingly reliant on high-end services. Then consider what happens when emergent companies layer in VoIP over Wi-Fi. Since VoIP translates voice communication into data, companies could charge a single flat rate that covers unlimited Web surfing, file downloading, e-mail and phone calls. But wait, the disruption is just beginning. What happens if companies go beyond charging a low flat rate that undercuts wireless carriers? What if they actually give away wireless voice and data service?
Now, before I move on from all of this theoretical stuff, some readers must be thinking “doesn’t this idiot know that muni Wi-Fi is in trouble; didn’t he read that EarthLink is backing away from this business?” Sure, I follow all the blow-by blow developments in the industry, but it is a mistake to assume this industry is in the fourth quarter, and its down by 17 points with 45 second to go. It is in fact early in the first quarter. How many of us remember that the apparent runaway success of the Apple iPhone was built on layers – on years - of miserable failure, beginning with the Apple Newton? EarthLink has certainly been an interesting player in this early market, but it is a mistake to assume that they define it, or that their decisions will determine its outcome.
Part of the reason any fledgling industry like muni Wi-Fi seems to always be on a roller coaster ride is that we read too much into press articles. The technology press is largely in the business of pumping up markets and technology to drive readership from early hype, and then tearing them down once the hype wears off, which drives yet more readership. The lesson for the industry is to stop knee-jerking and hanging on every slight shift in direction.
Now, I’ll turn to the real market dynamics in the mobile operator space. Why would Wi-Fi matter to a mobile operator? Is it a threat; an opportunity; neither?
Most mobile operators have made investments in Wi-Fi deployments, but these deployments have largely been limited to venue-based hot-spots. This seems logical; it’s a way to apply Wi-Fi for its intended use; un-tethered local area networking for nomadic users.
But when Wi-Fi started to cross over into the new domain of metro-area networks in the early part of this decade, I imagine debates started occurring throughout these organizations between engineers, product managers and ultimately executives. These are smart companies, with smart people. And Christensen’s books on disruptive innovation have always been standard reading in large technology companies. The product managers looking at emerging markets must have had a lot of fun in meetings with engineers who scoffed at this “crappy little LAN technology,” with its pico-cell architecture and its unlicensed (queue ominous music as Glenn Fleishman would say) spectrum.
So, despite the fact that competitive service providers, fledgling wireless ISPs, large, diversified incumbent providers and even some cable MSOs have been building, or in some cases experimenting with, muni Wi-Fi, mobile operators have stayed on the sidelines. Even T-Mobile, with its great Wi-Fi brand in the U.S., and its anemic 3G spectrum situation, has stayed out of citywide Wi-Fi RFPs. I believe there are five reasons that mobile operators will have to make their move in the next 3-6 months.
Reason 1 - Simple economics. The raw bandwidth and capacity delivered by hundreds of Wi-Fi radios deployed in a metropolitan area can be delivered at a fraction – a tiny fraction – of the capex and opex of a traditional WiMAX or 3G network. The fact that no spectrum costs have to be recovered is one reason. The high cost of cellular backhaul carried by mobile operators is another.
The kinds of rich content and media being downloaded and uploaded by consumers now and in the future will demand much more capacity than today’s mobile operators can deliver using their scarce spectrum and relatively large cell sizes. So, this is not about Wi-Fi as a technology; it’s about a fundamentally different architecture for wireless broadband networks; a pico-cell architecture. By placing radios closer to subscribers, the roughly 60 Mhz of Wi-Fi spectrum can be used and re-used over and over and over, in increasingly smaller geographic areas. Recent operator and investor interest in “femtocells” seem to validate that the market is going in this direction with our without Wi-Fi.
While the tech-minded critics of Wi-Fi complain about its low output power restrictions, the tables get turned when low output power is precisely what’s needed to re-use spectrum more efficiently and drive raw bandwidth and capacity. The old model of macro-cells, re-using 12-20 Mhz of licensed spectrum over several square miles won’t be able to stand up to this attack on raw bandwidth and capacity.
It should be noted that mounting assets (street light poles) can and will likely be a limitation with this pico-cell model. These assets simply don’t exist in a uniform way in all areas where an operator may want to deploy them. Not a big deal, since the mobile operator could simply use their WiMAX or 3G network as a “fallback” in these cases.
Reason 2 – Mobile and nomadic VoIP. While early muni Wi-Fi operators evaluated markets based on metrics like “household density” signaling that the business models were focused on providing a fixed-line alternative to DSL, the real opportunity is in mobile and nomadic services, particularly voice. Here is where mobile operators have the most to lose, and Wi-Fi has the most to offer. At the VON conference in March of this year, one of the panelists – a representative now working for Truphone – said he’d analyzed 20 cellular networks and found that “the average number of cell base stations used in the course of a call is just 1.06.” While I can’t speak to how accurate this is, what I can say is that it suggests that the vast majority of calls made on “mobile networks” are made when the subscriber is not in motion. In this context, mobile doesn’t mean mobile; it means not at home.
This means that one of the key value propositions (and barriers to competitive entry) that mobile operators have - seamless handoff and mobility - may not be the barrier to entry it has always seemed. If a large-scale muni Wi-Fi network were built over the voice calling area of an incumbent mobile operator, and services like Vonage, Skype and others were introduced in a big way, there would inevitably be a cannibalization of mobile voice revenues over time. This won’t happen overnight, as Skype didn’t destroy fixed line telephony all at once; and cellular voice didn’t replace fixed line telephony overnight; but it is exactly the kind of “chipping away” that can result in major disruption.
Reason 3 – Dual mode handsets and fixed-mobile convergence (FMC.) We know that there are several hundred million Wi-Fi enabled devices on the market already; everything from laptops to PDAs to portable game consoles to cameras. But Wi-Fi interfaces shifting into mobile/cellular handsets are coming, and coming fast . The recent launch of the iPhone has put virtually all handset makers on notice that they have to match the iPhone feature by feature, and that will include Wi-Fi. In addition, the rules set this week for the upcoming 700 MHz spectrum auction incorporate conditions that many believe will break up the strangle-hold mobile operators have placed on consumers’ choice of devices, once again leading to less ability for dual-mode handsets to be “throttled.”
On FMC, T-Mobile has been the most aggressive carrier to deploy Unlicensed Mobile Access (UMA) as a solution to integrate voice services between residential Wi-Fi networks and their cellular network. The larger operators have talked about UMA and/or IMS for years, but have dragged their feet compared to T-Mobile and many European mobile operators. T-Mobile is the smartest U.S. operator here, realizing that Wi-Fi is inevitable (at least in the home and office) and providing a service that integrates with Wi-Fi, particularly when they can charge for this integration, is better than the alternative of having voice revenues cannibalized by users who figure out that Skype on a handset is not so hard to use. One could also speculate that T-Mobile’s previous 3G spectrum constraints in the U.S., at least relative to the larger operators, have given them more motivation to offload expensive data traffic to Wi-Fi networks.
Reason 4 – Flat-rate pricing. Many have speculated that, if mobile operators dramatically reduced their 3G data prices – or move to flat-rate voice pricing – their networks could struggle under the demand. I don’t know if this is true, but the current $60-80 per month prices for unlimited 3G data would seem to be at risk if muni Wi-Fi networks spread far beyond where they are today. Due to the sheer bandwidth and capacity that Wi-Fi delivers, 3G (and I would argue WiMAX) will never be able to compete on a price per Mbps basis if a single mobile operator “breaks ranks” with flat-rate pricing. Sprint Nextel has already launched flat-rate pricing trials in some markets, so I don’t think this is far-fetched.
Reason 5 – Anchor tenancy. Capital-constrained muni Wi-Fi operators have shifted their business models to require revenue commitments from local governments recently, as a condition for agreeing to build Wi-Fi networks, and to meet cities’ social and economic goals. This has used by some to suggest that cities will pull back on their ambitious goals for citywide Wi-Fi, and that the muni Wi-Fi market will come to a crawl.
What these knee-jerk critics fail to realize is that the idea of anchor tenancy may be viewed quite differently by large, diversified incumbent providers than it is viewed by the less-diversified, smaller players. What I mean is that the ability for local governments to aggregate demand across a broader portfolio of services (wired and wireless, fixed and mobile) could be a windfall for mobile operators who step into this vacuum. Today, it may not be possible for local governments to aggregate enough anchor tenancy - using Wi-Fi alone - to get the kind of community benefit cities like Philadelphia, Minneapolis, Riverside, Houston and others got from these partnerships in the past. But when T-1 lines, cellular phones, hosting services and dozens of other portfolio items – which governments require anyway to run their operations – are introduced as a contribution towards this anchor tenancy… New opportunities emerge for public-private partnerships.
In summary, mobile operators face a Wi-Fi imperative. With its low barriers to entry and constant and continued innovation – and even with all its warts and blemishes - Wi-Fi is inevitably moving into their territory. Wi-Fi is the manifest destiny of Ethernet, without the cord or the wall-socket.
A classic “beat them or join them” decision is emerging, where mobile operators have to consider and analyze the risk of 1) doing nothing and hoping Wi-Fi just crawls back into the coffee shop, 2) deploying muni Wi-Fi aggressively as part of an integrated FMC strategy or 3) adopting some sort of hybrid approach. In this context, mobile operators stand where Digital Equipment Corporation (DEC) stood in the early 1980s. DEC stared at the fledgling PC market and its founder and CEO Ken Olsen famously said “There is no reason for any individual to have a computer in their home.”
We know where one company stands; as Verizon CEO Ivan Seidenberg was quoted as saying “Municipal Wi-Fi is one of the dumbest ideas I've ever heard.” I wonder what the other mobile operators will say.
Posted by Greg at 12:30 PM | Comments (0)