Strip-mining the radio spectrum
New connectivity technologies hope to extract the maximum value from a finite resource using low-power, long-range networks. But where are the standards, asks M2M Now’s new editor, Alun Lewis?
It’s always fascinating watching evolution at work in our industry. Just as out in the real biological world, as soon as new gaps, landscapes and opportunities emerge, so too do new technologies pop up to support and colonise them. Over the last couple of years, while cellular and satellite connectivity services for M2M/IoT have continued to develop, a flurry of activity has been taking place in another nascent sector: lowpower, long-range wireless networking.
Already, a number of technologies, vendors, communities and operators have appeared to fill this space – such as SIGFOX, LoRa, OnRamp, Link Labs, Weightless, SemtechNWave , M2Communication and many others. Their intention is to exploit advances in radio and other technologies to create networks specifically designed for IoT applications and devices – covering large areas, usually in unlicensed spectrum, and with the minimal power demands needed to deliver the long battery life required to support connected devices and sensors.
It is, however, early days yet. As a result, there’s still a lot of uncertainty about how these different approaches and technologies will compete and coexist – and which might eventually evolve into formal standards. This is a journey familiar to anyone who has been in the industry for any length of time and factors other than just sheer technological excellence will inevitably influence the paths eventually chosen. As Otto von Bismarck, founder of the German state, once famously remarked, “Laws are like sausages, it is better not to see them being made.”
Or, as one experienced industry commentator has sceptically observed about this specific space, “Some ‘standards’ are often just corporate objectives thinly disguised in sheep’s clothing when you look at them closely”.
Moving beyond cellular and satellite connectivity: first find your frequency
Michael Barkway, consultant at The Technology Partnership, sees the growing attraction of the “sub-GHz” domain, exploiting lower frequency ISM bands that are available around the world: “The promise of free spectrum is proving seductive for new entrants like SigFox, LoRa, OnRamp, and Coronis where a relatively simple protocol allows wide area – though contended – access. Rules and spectrum allocations vary across the world, but these lower frequencies gets inside buildings better and the use of “ultra-narrowband” helps manage signal-to-noise problems at long range, allowing very low-power devices. The downside is ultra-low data rates, and even further restricted downlink performance. SigFox, for example, is intended as a wide-area “public” network, with the other technologies available as either public or private networks. Bandwidth restrictions cause some tough challenges for software updates and security, but nonetheless, the technology is an interesting alternative for long-range, low-power deployments.
As far as standards paths are concerned, Barkway questions, “Will these technologies converge into one dominant standard for Wide Area Networks? At the moment, that’s not looking very likely. The market is in an early development phase now, with emerging standards competing with existing and nonstandard devices. Standards take a lot of industry collaboration, time and energy to produce, and there needs to be a strong and clear driver. That said, while technologies like cellular, WiFi, and Bluetooth required interoperability from the start, for M2M that’s less critical and non-standard products and WAN services have already escaped onto the market. Standards can however evolve by a more de facto approach too, and this looks like the more likely outcome in a substantial part of remote M2M access infrastructures.”
The role of unlicensed spectrum is also highlighted by Chris Howarth, head of networks for Accenture in the UK and Ireland: “Wide area M2M applications currently rely on a variety of networks such as WiFi mesh network, 2G mobile networks and proprietary networks such as Weightless, SIGFOX, Flexnet and OnRamp. In future, M2M could also rely on low power LTE-M, and choice of network will be dictated by M2M application requirements. The majority of wide area M2M applications however use very small data despite large numbers of transactions. The spectrum demand for M2M applications is therefore low compared with the anticipated demand for spectrum for mobile video consumption on smart phones and tablets. Under 1GHz spectrum is ideal for M2M applications due to low power requirements for M2M devices. Some European countries are currently trialling white spaces in the UHF band for potential M2M applications.”
“That said, Howarth adds, “Interference can be an issue in unlicensed spectrum bands if not coordinated well. In future, dynamic spectrum access could solve interference problems. However, the majority of M2M devices/ terminals would cost less than $2, so may not have intelligence built in for adaptive spectrum usage. Therefore, growth of the M2M sector depends on regulatory certainty on frequency band usage, and the monitoring of spectrum usage to identify potential interference problems. Frequency harmonisation across Europe for M2M spectrum would also fuel growth in M2M applications as it would provide economies scale for deployments.”
Readers can expect a significant acceleration of debate about spectrum issues in the coming months. With the next World Radiocommunication Conference (WRC) scheduled for Geneva in November, realpolitik issues of spectrum allocation will increasingly come to the fore as the many different players involved try to find harmony, while inevitably also protecting their own national or business interests.
Cast in concrete – or silicon – or a more dynamic approach as the market evolves?
In this context, the growing role of the Dynamic Spectrum Alliance (DSA) could be important for the M2M/IoT community, especially in developing countries. As it’s executive director, Professor H. Nwana, ex-group head of spectrum policy at the UK’s OFCOM regulatory body, explains, while the explosive growth of mobile data is putting enough strain on radio capacity, M2M/IoT demand will increase this further, driving a need for dynamic spectrum access to exploit both licensed and unlicensed bands.
“Dynamic spectrum access refers to a set of technologies which allow users to share access to spectrum,” Nwana says. “This will enable regulators to tailor technical conditions to the particular location and times where sharing is to occur; maximise spectrum usage efficiency and facilitate innovation; and protect incumbent services from harmful interference.”
In terms of the M2M community’s views on low-power, longrange connectivity, the International M2M Council (IMC) – a trade group that has amassed over 5,000 associate members in the past year, including both OEMs and enterprise users interested in deploying IoT solutions – intends to put questions about low-throughput networks to its membership in a new survey. “We definitely do not know what the awareness levels are about low-throughput network connectivity among M2M adopters at this time. We expect to see some differences between OEMs that are actually designing devices and enterprise users that are deploying them. Then again, I’ve been surprised by our group before,” says IMC executive director Keith Kreisher. The survey will be one of the broadest ever conducted in the M2M sector, polling IMC Adopter Members (associates) from at least six crucial vertical markets on five continents.
The technologies – coopetition, competition or peaceful coexistence?
As always where possible standards battles are concerned, there’s always strength in numbers and the low-power, longrange landscape is no different. So far, two main groups have emerged, each championing their solutions – the LoRa Alliance and the Weightless SIG – to sit alongside individual vendors championing their own solutions from a variety of perspectives. Formed at the start of this year, the LoRa Alliance already includes IoT solution providers such as Actility, Cisco, Eolane, IBM, Kerlink, IMST, MultiTech, Sagemcom, Semtech, and Microchip Technology as well as some major telecom operators including Bouygues Telecom, KPN, SingTel, Proximus, Swisscom, and FastNet – part of Telkom South Africa.
According to the Alliance’s acting chair, Staale Pettersen, “A LoRaWAN network architecture is typically defined as a starof- stars topology in which gateways act as a transparent bridge relaying messages between end-devices and a central network server. All end-point communication is bi-directional, allowing user downlink messages even in ad-hoc applications, while enabling network management and authentication. To support Over the Air software upgrades as well as mass distribution of messages, a multicast/broad cast mode can also be used.”
When it comes to communication between end-devices and gateways, Pettersen explains, “This is spread out on different frequency channels (FHSS) and uses different data rates, making a trade-off between communication range and message duration. With its spread spectrum technology, communications with different data rates don’t interfere with each other and a set of “virtual” channels is created, increasing the capacity of the gateway as well as reducing the self-generated interference and the probability of collisions. LoRaWAN data rates range from 0.3 kbps to 50 kbps. To maximise both battery life of the end-devices and overall network capacity, the LoRaWAN network server manages the data rate and RF output for each end-device individually by means of an adaptive data rate (ADR) scheme.”
When is a standard open or closed ?
For Professor William Webb, CEO of the Weightless SIG which recently launched the Weightless-N standard for lowpower, long-range applications, decisions about appropriate technologies must take into account much more strategic issues than just bandwidth and coverage.
He poses the question: “Imagine you’re a washing machine manufacturer in China and want to connect your device to provide a range of consumer benefits. What connectivity solution will you build in? Bluetooth, WiFi, PLC, 2G, 3G, 4G, proprietary SIGFOX or Semtech – or all of these in the hope one might work? Alternatively, imagine you make fitness gear and want to connect your device, then you’ll immediately reach for a Bluetooth chipset. That’s why we have a lot more connected wearables than connected white goods – and we need to do for the IoT what Bluetooth did for personal connectivity.”
“Firstly, it has to be a standard,” Webb suggests. “The washing machine problem shows that proprietary solutions just won’t work – and history tells us this too. Secondly, it has to be custom-designed for IoT, as cellular and WiFi over-deliver on data rates at the expense of cost and battery life. Thirdly, it has to be deployable by a network operator to provide ubiquitous coverage that multiple applications can leverage.
“While our Weightless-N technology is technically very similar to proprietary solutions from SIGFOX and others, it is a global open standard with royalty-free licensing,” he emphasises. “An open standard is much more than a notional label to describe a proposition with perceived traction in the marketplace – even a verifiably leading technology is not a ‘standard’ in the context of the relationship between supplier and developer. It is something specific; it defines the way in which developers, in this case of IoT products, ultimately pay for the connectivity technology. With a proprietary technology developers are locked into a single vendor and a single network operator. And monopolies never benefit the buyer. In an open standard, multiple vendors ensure competitive products both in terms of continuous improvement and in cost.”
NWave Technologies, an IoT connectivity technology vendor and core member of the Weightless SIG, is already heading up the Weightless-N working group developing the specification. Jonathan Wiggin, CEO of NWave explains, “Although originally founded in Russia, we’ve now moved our HQ to London, opened an office in Dallas in the USA and have a number of networks, both built by ourselves, or through partners such as BT in the UK. We firmly believe that proprietary technologies often end up supporting vertical silos as well as creating barriers to entry for application developers – whereas we go in royalty-free. The Weightless-N standard is rapidly evolving and we are on target to deliver compliant development kit hardware to IoT engineers in Q2 2015”.
Contrasting viewpoints and the rush to market
Meanwhile, SIGFOX has been busy, aggressively building out coverage for its low-power, wide range technology through a series of deals. Thomas Nicholls, head of marketing and communication at SIGFOX sets the scene: “Through the SIGFOX Network Operator partnership programme, more than 2 million square kilometres and 8 million booked subscriptions are already covered in partnership with local network operators. It’s already up and running in France, as well as the U.K., Spain and the Netherlands. It will soon begin deployment in the U.S. and other European countries. In addition, it is deployed in several cities, including Munich, Milan, Dublin, Lisbon, and Graz, Austria. SIGFOX Ready device makers can choose from a wide range of chipmakers, including Texas Instruments , Atmel, AXSEM and Silicon Labs.”
While standards issues will no doubt continue to generate a lot of dissent and discussion as each community and company continues to promote their own individual vision, it is significant that European standards body ETSI announced the formation in autumn last year of a standardisation group dedicated specifically to Low Throughput Networks (LTN) technology, aiming to allow object connections for just a few euros per year, with a few milliwatts for transmission and a modem costing less than one euro. The group has just released the first three specifications of an Internet of Things (IoT) network dedicated to low throughput communications. With potentially billions of devices just waiting to be connected – or so the pundits tell us – long-range, low-power connectivity is going to be one of the most interesting M2M/IoT enabling technologies for some time to come.