Over the past few quarters there’s been lots of talk about alternative wireless technologies to cellular and satellite communications, and the chatter has escalated to the point where the International M2M Council recently asked me to do some digging and comment on it, especially on the heels of some very recent moves in the space. Coming 100% from the cellular side of the business these technologies both intrigue and excited me as possible (and likely) answers to a few basic problems that we have in the M2M world. First is the morphing of M2M into the IoT with the idea that you can instrument almost anything if you can get to a very low cost module with many years of battery life and almost no access cost. This is one of the very basic underlying principals of IoT and the realization of the ’50 Billion Devices Dream’ (or John Chamber’s 54 Trillion Dollar dream, I’ll have some of that, thanks). The second is the shift in cellular access technologies away from 2G and even 3G towards 4 and 5G. Besides the fact that it just doesn’t make sense to use a broadband module for something that transmits just bytes of data (akin to using a $20K Cisco router at your home instead of a cable modem) there is the very real problem of 2G Sunset, and the uncertainty of 3G lifespan much longer after that, and not just in the USA (though clearly the US has the largest looming problem today) but eventually becoming a global phenomenon of history repeating itself, (like the move from analog cellular to digital) which we all know it often does.
I guess we should sidetrack a bit on why a 2G and 3G sunset in the first place? In the most simplistic terms, it’s not just about inefficient utilization of spectrum (though this is the biggest argument as it gets to the heart of matter), it’s also one of the simple economics of maintaining the legacy networks themselves. From some carriers what we’re starting to see is a general malaise around investment in the old gear. So if a 2G tower goes down in an area where the subs are predominantly 3G and LTE, don’t hold your breath for it to get fixed, you might just suffocate. For many operators, the 2G sunset is on the horizon and there’s not much we can do about it. On the other hand, there are MNOs who are ‘guaranteeing’ to keep the technologies alive into some distant (or not too distant) future, but one wonders what a guarantee actually means in the midst of M&A deals and corporate re-orgs.
This is the uncertainty that we’re dealing with in the M2M business and in some market segments it’s a real pain. Again, though this is mainly a North American phenomenon at the minute, the concern is global, and business models that require 7-10+ years of equipment lifecycle are truly at risk of being stuck in a technological cross current, so it’s no joke.
Therefore the general consensus is that one better start thinking about LTE or alternatives to cellular for wide area access, and the non cellular options kick in especially in the low data rate and low power environments while LTE has already become the de facto standard for more feature rich applications like connected cars and kiosks. An application-cloud developer friend of mine recently told me that over 50% of his projects today are non cellular, mostly WiFi at the moment, and that figure is growing, and there appears to a whole new age of IoT’ers who have seemed to forgotten that cellular even exists. So eschewing cellular for in-building or campus environments is becoming more and more prevalent.
But of course many applications need long-range wireless connectivity and the cost of LTE today is just not an option, and quite possibly never will be. So that’s where what some call “Sub Gig” comes into play: the many flavors of Low Powered Wide Area Networks (LPWAN), plus other technologies like White Space (which utilizes empty channels in the TV Spectrum).
When I first started preparing to research and write this article I thought I’d have a few discussions with some industry experts, get some answers, and go merrily off and put finger to iPad. However, each generous discussion led to more questions and more information about different technologies and their intricacies. And so what I thought would be a short primer on LPWAN technologies over unlicensed ISM bands has become a pretty comprehensive yet broad based overview as to what’s out there.
First of all, there are numerous companies that are operationalizing various forms of LPWAN (I’ve counted nearly ten) in various forms, mainly categorized by super low bandwidth, very low operating costs (less than a dollar per month, in some cases less than a few dollars a year, or approaching free) and modules/chips as low as $2. We can break these down into a few sub-segments:
– Open systems under the Weightless Alliance.
– Proprietary systems under the LoRa Alliance, using Semtech chipsets.
– Proprietary systems that are part of no alliance, such as SigFox.
– Systems using truly open architecture (like 900 MHz).
Then there are also potential rival lower band cellular technologies that are meant to make LTE less costly (called LTE MTS, LTE M, LTE Cat 0) but as of today these appear to be quite a ways off from being operationalized, possibly years. This is what Neul, recently acquired by mobile technology giant Huawei, is now developing after seemingly abandoning white space. More on that later.
And so, there are three schools of thought here: one is that Mobile Operators are the best equipped to operate standards based networks for IoT, and that they should be subsets of their existing spectrum, or possibly new spectrum will be acquired specifically to operate low band IoT, hence LTE Cat 0. The second is that a new breed of network operators will exist specifically aimed for operating IoT, but that they should do so under industry accepted standards, arguing that non standard technologies over time prove to just not win, and standard ones do (like Bluetooth, Wifi, Cellular). And the third are companies that appear to be closely guarding their technology as to not spur competition in the market, and instead build on proprietary technologies, for better or for worse. In all cases, taking recent announcement into account, the MNOs are getting involved, and in a big way.
With respect to the LPWAN players, the good news is that all of them appear to be viable in the specific applications that require very low bandwidth and in some cases uni-directional (one way) communications.
The bad news is that it’s a little early to know how things are going to shake out and that includes with many of the companies building business models around the technology. What you are actually able to do with the freely accessible ISM bands depends very much on the regulatory environment of the country you are working in, parts of Europe are a little more complicated than the USA for example, and exactly which bands are available for use and how you can access them vary regionally. These regulations have to do with Duty Cycle, basically how often your devices can be actively pinging the network, how much data they can send, and over what range. This affects the battery life of the device and the cost of the chipset. The quality of service and the security are affected as well.
Of course, there are some very smart guys employed by all of these companies figuring out how to deploy a viable service under such restrictions, and they appear to be doing it very well.
So what is the differentiator between these technologies? Pretty straightforward it seems:
Weightless is an alliance, not a technology, and at least one ISM Band player has joined, nWave. nWave are already operating in the UK in a very high profile project in Milton Keynes with BT and are making plans to deploy services regionally in Europe and also in the USA and they have ongoing pockets of business and trials in both markets. Other industry heavyweights who have joined Weightless include Accenture and ARM.
The technology backed by Weightless is unidirectional at the moment, but future revisions in the coming quarter plan bidirectional service, and are only suitable for ultra low band applications like sensor networks (such as smart agriculture, smart parking, smart grid and smart cities).
LoRa is an alliance that has gathered around a single proprietary technology developed by chipmaker Semtech. Companies in this category include Kerlink, Senet, and long time M2M Equipment manufacturer Multitech Systems. Startup Senet is deploying services in the USA, and others are popping up in global projects. Most notably Dutch Mobile Operator KPN recently announced that they were joining LoRa and are planning to deploy the technology in Holland, and a slew of other MNOs have jumped on board recently as well, including a fully operational network in South Africa run by Fastnet, an M2M-focused subsidiary of South African’s Telekom MNO.
In the USA at least, the technology is bi-directional and is already being used widely by Senet in tank monitoring applications nationwide and appears to gather traction. But the way in which LoRa might do bi-directional in the European market is fuzzier because of bandwidth restrictions on duty cycle.
US Equipment Manufacturer Multi Tech Systems has also announced a suite of products using LoRa technology that includes both programmable modules and Gateway/Hubs, also integrated with their legacy cellular hubs, all available in an April 2015 product launch which looks particularly promising.
In the proprietary non-standard camp, SigFox has been well backed by Intel ventures and others, and has been operational in France for some time, and pushing into other markets through partnerships. The most notable of these is the recent announcement by Arqiva that they will jointly pursue UK Metering with SigFox, and another announcement with Swedish MNO Tele2 that they will deploy SigFox technology in Holland. Then in the middle of writing this article, SigFox announced a whopping investment of nearly $114 Million by a group of investors including Telefonica, SK Telecom and NTT Docomo.
So we now have different MNOs backing several different camps, which is a good sign across the ecosystem.
Also recently yet another startup, Helium, have garnered lots of attention by raising its own boatload of money (backed by Sean Fanning of Napster and Facebook fame amongst others) but yet to announce any specific technology or real plans for deployment.
Going back to traditional cellular for scalable low bandwidth applications, LTE Cat 0 seems to be a ways off because of the time it will take to go through full 3GPP certification and for equipment vendors besides Huawei to jump on board (e.g. NSN, Alcatel, Ericsson, etc). It looks as if the potential to use LTE for low bandwidth, low cost modules is at least a year away from proof of concept, but it is anticipated that it will eventually happen.
The skeptics abound here. Dr. William Webb, the CEO of the Weightless alliance and a renowned industry expert on wireless technologies likens using LTE for IoT applications to using a Ferrari to do the job of a pickup truck. He also feels that the MNOs are more interested in 5G and, dare we say, 6G. We shall see!
There are also some really interesting technologies using 900MHz band, one just developed and operationalized by equipment maker Digi for instance, which is more of a point to point higher bandwidth solution with over 100 miles of range and a patented technology for throughput called Punch2.
But what about good old fashioned 2G? Boy do we love it…. works like a charm and getting cheaper by the minute. If a 2G module were under $5 and the network operator would just keep it going, why not? It does appear in certain markets that this will be the case, and in the USA Sprint is saying they will keep their 2G around well into the ‘20s. So will they?
At least one company has started up in the USA with a concept to build an M2M specific GPRS network using patched together spectrum: M2M Spectrum Networks. Seems promising if they pull it off, and maybe a blueprint for 2G longevity globally, however is there enough business in the future to support a dedicated 2G spectrum for IoT, especially with such low cost ISM based technologies available?
The fact is, one cannot argue the allure of close to free modules and access for IoT, yet concerns about scalability, security, regulations and standards do exist. And the trillion-dollar question is: Can the ISM bands really support a gazillion devices, especially with so many competing technologies in one small slice of spectrum? On paper they appear to do just that, but the devil is in the details. However, as these technologies mature there is no doubt that they will address this.
So what will win? The proliferation of the ISM band technologies for low band communications is actually nothing new, it’s just getting more reliable… the world is spinning faster and faster and new applications of old technologies are popping up all the time. For all we know, in 10 years we’ll be travelling around the USA in air tubes owned and operated by Elon Musk.
My new friend and industry analyst/heavyweight James Brehm puts it best: ‘Right tool for the right job.’ Well-said James. Whether it’s LPWAN, Cellular 2g, 3g, 4g, Satellite, Wifi, Zigbee, mesh networks, or two tin cans and some string, in this brave new world of connected everything we now have a ton of tools at our disposal and there is no doubt, as is already evidenced, that smart people will find new ways of making use of them, the question for me, as always, comes down to monetization and ROI.
But confusion does abound. Webb likens this to what he calls the ‘Washing Machine test’. Imagine today that you are a manufacturer of Washing Machines and you have decided for some reason to layer in connectivity, I don’t know, for soap as a service or something. What technology do you put in there? Realistically today the answer would have to be wifi? (That’s a question). Or like Bosch Security systems maybe you offer a plug in card that can be swapped for new technology in the future. For them today 2G and 3G cellular and tomorrow likely LTE.
What to do….?
Use the Right tool for the job! Thanks James.
Special thanks to all who gave of their time to be interviewed for this article:
Matthew Bailey of nWave
James Brehm of James Brehm and Company
George Dannecker of Senet
Paul Egan, former VP of Engineering of Neul
Tracy Hopkins of Stream Communications
Aaron McNally of Digi
Chris Rezendes of Inex Advisors
Duane Wald of Multi Tech Systems
Dr. William Webb of the Weightless Alliance
Comments and Questions welcome.
Opinions are my own and 100% accuracy is questionable!