It should be simple right? Choosing a wireless technology to port small amounts of data from A to B. Not very fast, and not very often. How hard can it be? After all, we’ve had technologies that allow us to instantly stream movies in HD to portable devices in almost every corner of the world for a few years now – this should be super easy. So as we explore the landscape you can be sure that we’ll quickly home in on the requirements and identify the obvious candidates. Good!
Not so fast. Making sense of the wireless connectivity technology landscape for the Internet of Things is the opposite of simple and obvious – it is a complex minefield and getting more complex with every week that passes. So before we attempt to get our arms around this picture, let’s take a moment to review the main suspects – a quick scan of just some of the various technologies vying for a piece of the IoT cake. This is not an attempt to list every contender in the race, we’re just establishing a framework for the discussion and I’m going to describe an arguably simplistic and arbitrary way of segmenting the market – short range vs long range. This makes sense because, traditionally, short range technologies have generally tended to exhibit low cost and low power/energy consumption whilst long range technologies occupy the other end of the spectrum – high cost and high power. Cost, power and range are three of the fundamental characteristics central to any IoT connectivity tech so this should help. If but it was that simple. What the majority of IoT use cases demand is a hybrid of the two extremes which is the driver behind the convergence of a raft of alternatives. Our utopian Shangri-La is low power, low cost and long range and as we’ll see, this is where the market place is starting to get very crowded. A useful analysis of any spectrum might perhaps start with the extremes so that’s what we’ll do now.
In the short range corner we have some familiar LAN and PAN technologies – Bluetooth, and in particular Ble, Wi-Fi and even Zigbee. For sure there’s a bright future for these options, particularly Bluetooth low energy – depending on the use case. Local area network technology will find IoT applications in the home and office where range is not a priority.
Over in the long range corner we have the familiar 3GPP. Or perhaps the current ‘alphabet spaghetti’ of terminology in 3GPP world is no longer so familiar. If you were reading this just five or so years ago it’s almost certain that the discussion would be about, at best, novel alternatives to the ‘mainstream’ 3GPP option. Non-cellular technologies certainly weren’t regarded as a legitimate segment and solutions from the 3GPP camp were widely seen as the natural favourite in the race to connect up the billions of things that are coming.
Well the world gets real complex real quick. In 2015 the GSMA were heading, so they thought, so they hoped, or so they told us, towards consensus. An embryonic 3GPP based IoT standard was emerging and we could look forward to it at the end of the year. As it evolved and metamorphosed through various iterations we tracked the progress of the nomenclature if not the technology. So far we have enjoyed LTE-M, LTE-MTC, Cat 1, Cat 0, Cat M, Release-12, Release-13, CIoT, Extended Coverage IoT and NB-IoT and the milestone date in our diary has slipped to March 2016. So what can we expect then?
NB-IoT is the long term goal of the long term evolution bandwagon. What we have until that arrives is Cat 1. In between, Cat 0 will be skipped outright, and we will also have Cat M as a stepping stone.
The difference between Cat M / NB-IoT and Cat 1 is in the power consumption and cost. Cat 1 is not where it needs to be on either – an optimised IoT connectivity technology it is not. It is the interim solution until we get to Cat M / NB-IoT. So when will these actually arrive? As I write this, March 2016 is a matter of weeks away so let’s dispense right now with that fiction. It’s not two weeks, or two months. Cat-M is scheduled to be commercially available in 2017, while NB-IoT is announced for 2018. But none of it is here yet.
The cellular industry took its eye off the IoT ball for a few years and that created a gap that companies with names increasingly familiar to us are now fighting over. The cellular community wants to reclaim this lost ground quickly before the pretenders become too deeply entrenched but will it succeed? Well it’s tough to do when a technically competitive 3GPP IoT solution is not even close to market ready. Expect a ferocious media bun fight in the interim as the GSMA reassures the market of NB-IoT’s imminent arrival.
Ultimately it won’t be a single race of course, and there won’t be a single winner. NB-IoT or whatever it is called when it finally arrives will undoubtedly play a very significant part of the IoT landscape but not yet and then only for use cases where there is a good fit. Applications where the value of the individual data streams are relatively high and where energy consumption is not so critical are a natural fit. The reality of required backward compatibility with a complex legacy (LTE) technology is that no amount of optimisation will be competitive with a true ‘clean-slate’. Which is where I will now turn.
The simplistic LAN vs 3GPP view from a decade ago is not even close to realistic today and the dozen or so propositions circling the middle ground makes this a complex industry to analyse. Low power, wide area network technologies (LPWAN) in sub-GHz, licence exempt spectrum have surprised many by arriving, as if from nowhere, to be pretenders to the IoT connectivity crown. LPWAN is the black fly in the chardonnay for the 3GPP community and far from an interim solution until NB-IoT shows up, it is looking increasingly established. Perhaps the real battle is within the LPWAN segment so let’s briefly review the players.
In 2009 French start-up, Sigfox emerged with an unsophisticated ultra narrow band modulation scheme and an operator rather than a technology model. The company has since become one of the best known vendors in the current LPWAN market. In the process it has arguably done more than any other company to evangelise the LPWAN segment with aggressive marketing, solid finance and an enthusiastic, rapid fire PR department generating awareness. Around that same time the Weightless SIG appeared, a global standards body aimed at democratising and defragmenting the LPWAN market with true, royalty free, open standard IP. The initial technology from UK based Neul, later re-branded as Weightless-W, was designed to operate in TV white space spectrum. Technically superb. A commercial failure. Dynamic access to white space spectrum for secondary users was highly regulated and simply did not open up as quickly as anticipated. Neul switched direction following a change in the management team at the the top of the company and its technology can now be found in a Huawei licensed spectrum proposition. Fast forward a couple of years and two more contenders appear. Semtech bought French long range wireless IP company, Cycleo, in 2012 and subsequently launched its wide band technology. Meanwhile, Russian company, Nwave, emerged on the market with another UNB technology and elected to make this open source through the Weightless SIG. Weightless-N was born. And in 2015, with, on paper at least, the most advanced technology in the LPWAN segment to date, Taiwanese company, M2COMM, announced a sophisticated narrow band technology claiming to bridge the gap between UNB and wide band modulation schemes to provide optimal performance for the majority of IoT use cases. This was subsequently to be developed into an open standard and offered through the Weightless SIG’s standard body as Weightless-P. On-ramp, subsequently renamed Ingenu, arrived with an interesting alternative to LPWAN, Random Phase Multiple Access, RPMA, technology claiming better capacity and range. Other players include Russian company, Strij, with what appears to be a remarkably similar solution to that of both Nwave and US company, WAVIoT. Cynet from Cyan, Qowisio, Starfish from SilverSpring, Accellus, Telensa and DART do nothing to simplify the complex LPWAN market with yet more offerings.
Well our mission is going to be to make sense of this raft of technologies – a non-trivial challenge! Our goal is not to forecast winners and losers – we don’t believe that the IoT market will polarise so digitally. Instead we will identify the use cases and generic types of technology that are most suited to them. There will be applications in the home and office environment that Bluetooth and Wi-Fi provide a perfect fit for. There will be high data value applications with commercially feasible connection to mains electricity that will almost certainly favour an NB-IoT technology and there will be a lot of power sensitive, low data value applications where LPWAN technology offers a compelling rationale. Our objective over the next few articles is going to be to identify the parameters that matter so that IoT developers can make informed decisions about the platforms on which their products are based. It’s going to be an exciting journey!