Is it better to suffer the slings and arrows of outrageous fortune by keeping inefficient 2G networks going in order to let customers run low capacity M2M devices at optimum cost? Or should operators rise up against them and force them over to more expensive 3G solutions that they will probably never need, and in so doing, risk facing a sea of troubles?
Of course, the real rub, as Hamlet would have said, revolves around the bill - and who is going to pick it up. It's one thing clearing the decks of unlovable 2G spectrum in order to create bandwidth for high-value 3G and 4G revenue generators. It's quite another expecting M2M customers to pay for the clean up.
Many M2M devices that are currently out there in the field are deeply embedded in remote locations. They can very often be expensive to get to and time-consuming to replace and they were initially installed in the expectation that they would have a very long shelf life in situ - anything up to 20 years.
Moreover, they typically use very little capacity, send data only sporadically and rely on very cheap chips to make them economically viable. In short, they do not represent the bandwidth-hungry segments that will drive M2M mobile data volumes skyrocketing in the future. But they do represent a long term, stable source of revenue that few operators will want to risk turning their backs on so early in the burgeoning M2M life cycle.
One of the biggest challenges facing M2M providers, therefore, is not how to develop new opportunities in the market place, but where to house those new customers on the network. Put bluntly, that means setting aside ultra scarce spectrum to support M2M communications when mobile operators are desperately trying to throw as much spare capacity as possible at the roll out of 4G.
Sprint takes a tentative step
This June, Sprint will shut down the push-to-talk 2G network it acquired as part of its troubled $36 billion merger with Nextel back in 2005. The move, which was first announced last year but has been under consideration since 2010, is part of the carrier's Network Vision initiative and is aimed at simplifying Sprint's portfolio. The ultimate goal is to free up enough spectrum to allow Sprint to run a 3G CDMA network as well as roll out a 4G LTE offering.
Analysts warn against reading too much into the IDen shutdown, as Sprint continues to run its own legacy 2G network. Meanwhile, Tom Nelson, a director of M2M marketing says that Sprint is engaging with its customers to offer a range of solutions to meet their needs: "We have very specific contracts with certain customers where longevity is extremely important. Clearly, they are very interested in our network vision and we will sit down and explain our road map to them under non-disclosure agreements. We haven't felt the need to go out there and pound our chests."
But as the death knell sounds for Sprint's IDen network, rivals will still want to scrutinise the churn rates. It is not clear how many subscriptions on Nextel's legacy 2G network are pegged to M2M devices but around the time that Sprint confirmed the move in May 2012, just one in three subscribers who terminated their contract with Nextel took up Sprint's offer to move over from 2G to 3G. The remaining two-thirds, found a new network provider altogether.
Those numbers will not make for comfortable reading over at AT&T. Last autumn, America's biggest mobile operator said it too would shutter its 2G network by 2017 as part of a root-and-branch upgrade. Around 8.3 million subscribers, or 12% of its contract wireless customers, will be affected by the move.
AT&T is adamant that it will manage the migration "proactively" and that the move will not impact on revenues in any way. The carrier is one of the biggest providers of M2M-based alarm systems in the country and industry specialists estimate that as many as 3 million customers could be affected by the move.
That leaves Verizon. Hard on the heels of AT&T's move, America's number two player said that it would stick by its 2G network until at least 2021 and, furthermore, would give M2M users at least a decade's warning before closing down a network.
Verizon is in a slightly better position than AT&T because its LTE roll out comes from new spectrum, so it doesn't have to find capacity from its existing inventory. And it clearly plans to strike home that advantage with a vengeance.
Round 1 in the public relations war, then, goes to Verizon.
Cheap as chips?
Holding potential M2M developers back from 3G is not some deep-rooted fear of new technology, but the high price of next-generation chip sets. On average, modules account for around 30-40% of the total cost of an M2M device.
As recently as two years ago the cost of a 2G module was around $20 while an equivalent 3G chip could easily command five times that number. Today, a device developer can produce a 2G chip set at scale for as little as $2.50, while a 3G module costs $40-60 and an all-new 4G chip costs $80- $100.
These economics present a very real conundrum to mobile operators: On the one hand, they need to migrate customers over to 3G connectivity - partly to clear spectrum so that they have enough room to support far more profitable bandwidth on 3G and 4G networks, and partly to up-sell customers value-added services that only 3G compatible devices can deliver. And yet on the other, they cannot afford to risk pricing themselves out of the market by limiting the potential customer base to those willing to embrace 3G solutions.
There are those who argue that M2M customers should not expect a choice of network operability in the first place.
The head of M2M at one North American carrier, speaking on condition of anonymity, argues it thus: "As far as we are concerned, the M2M market is not yet significant enough to play a role in determining how and when we move from 2G to 3G and 4G."
"While it is almost certainly true that most M2M applications would happily remain on 2G for the foreseeable future, the reality is that they are not in a position to choose. They will have to come along for the ride. Customers should not expect to stand in the way of progress."
The M2M chief accepts that the transition might prove difficult, but argues that such problems will be a fact of life for many M2M applications in the future anyway: "We support a lot of telemetry applications in the oil and gas sector. The devices are intended to operate out in the field for 20 years - but in reality network topologies will change in a fraction of that time. It would be ridiculous for a customer to expect a technology like 2G, which is already more than a decade old, to continue running out to 2035."
The issue, he warns, is not specific to the 2G-versus-3G debate, but will resurface regularly as networks continue to evolve. "You can speculate as to how useful over-the-air firmware upgrades are likely to prove in the future, but if the network topology changes, then I don't see any way round a rip and replace overhaul." Providers need to start a dialogue with customers in order to decide who should shoulder the cost of network upgrades and how they should be factored in to the M2M model, he says. "I think we have a way to go before we nail the problem", he adds forebodingly.
Others suggest that customers will baulk at migrating to a 3G solution: "Around 95-98% of M2M traffic is best served on a 2G network", says the head of global M2M services at one leading European mobile operator. "I can see a situation developing where M2M customers ultimately migrate from 2G to 4G but even then, I wouldn't bet the farm on it. As far as moving over to 3G in the interim goes, frankly, I just don't see the rationale. Customers don't need it, they don't want to make the migration and they can't afford the chips. That's a pretty damning case against the premise."
Two further options are worth exploring: Firstly, carriers could set a side a small chunk of spectrum to dedicate to 2G M2M services. According to KORE Telematics, a Canadian-based provider of wireless M2M services, every single M2M device in existence at the end of 2016 could be supported with just 5 mHz of spectrum on a 2G network.
But the problem, explains chief executive Alex Brisbourne, is that operators chasing average revenue per unit of $80 or more on a 4G smartphone, struggle to justify setting aside spectrum - however tiny - for an M2M device that might generate just $5. "The carrier community has no real incentive to keep 2G spectrum going – even a sliver as thin as 5 mHz."
Secondly, operates might try to develop a new network altogether dedicated to M2M on its own frequency. Such an option is laden with risk, as Lightsquared found out when it tried to wring new spectrum out of the crowded airwaves in the US.
Can baby monitors save the day?
But where angels fear to tread, a small French start-up called Sigfox is more than willing to go. The Toulouse-based company, which is backed by US chip giant Intel, claims to be close to rolling out a network architecture using unlicensed low frequencies commonly used by baby monitors and cordless phones.
Sigfox claims that it can achieve the same coverage with just one tower that would require up to 100 cell sites at a traditional mobile provider. It is currently rolling out a network across France using just 1,000 transmitters and says that it could do exactly the same in the US with as few as 10,000 towers.
The catch, is that Sigfox's network can only support bandwidth speeds of just 100 bps, making it up to 20 times slower than 2G. While that will never support video streaming and other value-added services that Big Telco is pursuing in the M2M arena, it is perfect for many solutions, such as security alarms, that only trigger in the event of an incident. And because Sigfox's network is cheap to run, the operator reckons that annual connection charges could be as low as $1 a device.
By its very nature, it is likely to prove particularly attractive to those very customers that are so reluctant to upgrade to 3G. Moreover, says Stein Soelberg, director of marketing at KORE, the Sigfox model is interesting because it flies in the face of suggestions that M2M connectivity is all about the proliferation of Big Data. "Our position has always been that Big Data is a misnomer – it is our contention that most devices will report on exceptions rather than merely relaying huge streams of information and that has implications when it comes to bandwidth requirements."
The scarcity of spectrum in the US is undoubtedly hastening the demise of 2G networks. In Europe, on the other hand, operators are more prosaic over its future. That might be partly because it is far more common to sign assurance of service contracts in Europe, and many of these have as much as a decade still to run. But it is also likely to reflect the huge investment that continues to go into 2G technologies.
"We have no plans to switch off any 2G networks," says Marc Saulter, head of global M2M business development at Vodafone. "At the moment we are still investing in our 2G business, so when we start working with a new customer we can instantly provide a commitment that we won't be shuttering our 2G networks any time soon." Saulter prefers not to be drawn on any specific timeframe, stressing that clients will be provided with appropriate guarantees on a case-by-case basis. "But typically, I would say that we are talking about no earlier than 2020. We don't see any necessity to switch off 2G - its still sufficient for most of our customer's needs."
Rodolphe Fruges, vice president of M2M at Orange Business Services is similarly sanguine: "We will maintain 2G until a minimum of 2020 - that's a good horizon for most customers. For now, the bandwidth requirements of most M2M developers are very low – and that is going to encourage them to stay on 2G networks."
By the time they face pressure to upgrade, the issue will not be whether to switch from 2G to 3G, but whether to make the jump straight to 4G, he says.
