Simple solutions create complex connectivity choices for IoT service providers
Too much choice can be a dangerous thing but, the vast diversity of IoT applications and business models depend on appropriate functionality being available at varying price points. George Malim examines the options.
LTE, 5G, Cat M1, Cat 1, satellite, low power wide area networks (LPWANs), narrowband IoT (NBIoT), Sigfox, Bluetooth, ethernet and more, the list of connectivity options for serving Internet of Things (IoT) applications and services appears to be almost endless. However, the decision about which to select is simplified by the nature of the market. Some technologies are simply too costly, too slow, too unreliable, too power hungry or just unavailable to support the needs of applications and their business models.
At the high end of the market place, satellite communications can be used to assure universal coverage but this is not suitable for a massmarket with continuous communications needs. LTE, and much later 5G, don’t have complete coverage and, while the security and bandwidth attributes are attractive, the cost isn’t.
This leaves providers of mass market, high volume, low value IoT services looking outside satellite and high-end cellular communications to find the right connectivity to support their offerings. A relatively new wave of options in low power radio and the lower reaches of the cellular range is emerging. Principal among these are three groups of technologies: LPWAN, NB-IoT and Wi-SUN. Each has its advantages, although NB-IoT and Wi-SUN are in their infancy.
The challenge therefore for users is to identify which technology most closely serves their customers’ needs and the goals of their businesses. “You cannot compare a set of usage of one company to the set of usage at another company in another industry,” says Christophe Fourtet, the founder and scientific director of Sigfox, an LPWAN technology with operations globally. “It’s a very long process to compare these technologies but we’ve been trying to accelerate it. Since the beginning of Sigfox we’ve had the same target of shooting for the massive low-cost IoT market. Instead of being focused on technology and performance, we have aimed more for an extremely simple device that costs very, very little so you can deploy massively.”
The decision also hangs on your role in the market place. Are you a user or a deployer? “There are two perspectives to consider, one relating to organisations deploying radio technologies and the other those adopting or using them,” says Ken Figueredo, an IoT strategy industry advisor to InterDigital. “The first category applies to network operators or connectivity service providers. Their legacy investments and technology roadmaps govern their decision process. In practice, connectivity service providers will see demand for hybrid solutions that combine different approaches due to the heterogeneity of end-user needs.”
The number of issues to be considering is significant. “Which technology to select depends on many questions,” acknowledges Phil Beecher, the chairman of the Wi-SUN Alliance. “What is the reliability you need? What security? What latency and do you need local control? should all be considered. Another consideration is the business model. Do you want to spend capital on equipment and manage the network yourself or do you prefer an opex model? That will decide organisations between LoRa, NB-IoT or third party LoRa.”
“One of the fundamental differences between WiSUN and the others is that Wi-SUN can be configured as a mesh network and it can also support faster data rates and lower latency,” Beecher explains. “LoRa and NB-IoT offer hub and spoke connectivity but we support mesh so you can have local control. Local devices can communicated with a local route so you can run local actuators. In addition, the higher data rates can also support security negotiation.”
Neal Forse, the chief executive of WND UK, the UK’s Sigfox network operator, urges caution: “Any organisation considering these competing radio technologies, first needs to evaluate the landscape as it exists today. NB-IoT is not yet commercially available, and will likely not be until 2018 and even later in the US. For businesses looking to design and deploy their devices and solutions in the here and now, this narrows the field of competition somewhat. Even when NBIoT is ready for prime time, it will take months, if not years, in order to reach the same level of maturity as some of the competing LPWAN technologies on offer.”
“Sigfox has first mover advantage in that there is already a wide range of successful deployments and the technology has a rapidly growing ecosystem of partners,” says Forse.
The decision process comes down to matching the technology to the organisation’s business drivers. “In the case of firms adopting and embedding connectivity technology into their products and services, the decision is not a purely technical one,” adds Figueredo, pointing out that other factors influence the process including:
- The business case Is there an acceptable return on investment (ROI)? What is the business risk from locking into a given standard and supplyside ecosystem?
- Market timing Is it important to get to market quickly, possibly using a proprietary approach, with an acceptable risk if the technology needs to be exchanged or upgraded at a later point in time?
- Capacity building What internal competencies will an organisation need to support connected product or service offerings without amassing significant investments and personnel or technology-partner risks?
Companies thinking about their IoT deployments could be forgiven for glancing at the technologies and concluding the offerings, aside from the technological debate, are largely similar and the decision should be simply one of selecting the technology that offers the best coverage at the lowest price.
“It’s definitely more complex than that,” explains Fourtet. “If you make a bet on a public network, and I believe that’s almost the only way for massive IoT, the resources that provide the service to your devices so the cost is as low as possible have to share infrastructure. It’s a mistake to think you can succeed with a private network outside of very high value scenarios.”
Figueredo acknowledges functional similarity but points to the differences in the details. “The basic functionality of different low-power, wide-area IoT radio technologies is similar – they support long service-life devices and low data rate connectivity at low price points compared to traditional mobile connectivity technologies,” he says. “Individual technologies, both proprietary and standards-based, vary in the way they work whether in terms of duty cycles, data transmission characteristics, upgrade capabilities or other factors. Operational characteristics could affect how a commercial service is delivered and should be one consideration in the technology selection and product and service design process.”
Forse refutes the argument that the technologies offer broadly the same functionality. “The solutions are in fact vastly different,” he says. “For a start, NB-IoT is a cellular technology, operating in the licensed spectrum, whereas Sigfox and LoRaWAN are not. Both cellular and non-cellular have their own unique value propositions and each can fulfil use cases that the other cannot. NB-IoT works – or will work, when it is marketready – best for applications that require minimal latency, high throughput or highly frequent communication. Technologies like Sigfox are ideally suited to applications that need to be delivered at a very low cost, have less frequent communication requirements, and require exceptional battery performance.”
There will probably be a harmonisation among the different types of LPWA providers that will ultimately make the market less bewildering to navigate but there is a long way to go in terms of brining clarity to the market place. “NB-IoT involves the costs of managed spectrum and the mobile network operators will have some control over the size and deployment of the network while, with LoRa you just don’t know how many nodes are going to be deployed and how, if they are densely packed, they will start interfering,” says Beecher. “Wi-SUN fits nicely in the middle of the segment because our lowest data rate is way above that of Sigfox and goes up to LTE Cat M. Wi-SUN still needs backhaul and some of that will be fibre.”
The differences in performance and other attributes between the technologies mean they are likely to co-exist, although the proponents of each are keen to enforce the advantages of their offerings.
“If it were the case of simply selecting a technology based on cost, Sigfox would win hands down each and every time,” says Forse. “Sigfox has the lowest cost radio modules, by a significant margin. If we look specifically at LoRaWAN versus Sigfox, there are some important differentiators here too. LoRa does not provide a network to its customers. The LoRa Alliance has developed the standard and the business model is focused on selling chips.”
“That means if you want to use LoRa for your devices, you either have to create your own network, managing the gateways and backend yourself or use a network operator that offers LoRaWAN networks,” he claims. “The former requires significantly more investment and expertise to instantiate and then manage the Lora gateways and the associated cloud backend, while the latter limits the geographical reach, as each LoRaWAN operator has a separate geographically isolated network; none of which share the same OSS (operations support system) or BSS (business support system) infrastructure.”
Figueredo thinks that the sheer scale of the IoT market means there’s room for everyone and organisation will value having the choice of solutions that more closely target their needs. “In the short-to-medium term, each of the different technologies will co-exist because the IoT market is massive,” he says. “As a result, there will be user segments that value different commercial, operational, supplier ecosystem and technical characteristics.”
“Over the medium to long term, we expect module price differentials to narrow such that scale and service capabilities dominate,” Figueredo adds. “It is therefore important for the designers of radio modules and the network operators or service providers to build up their service enabler capabilities to manage LPWA endpoints, to support new uses of these endpoints via functional enhancements, and to manage hybrid deployments that integrate more than one connectivity technology.”
Even competitors acknowledge the different value propositions of rival technologies. “The other technologies have a place,” agrees Beecher. “If you want ultra-low transmission power, rather than low energy usage, LoRA is very successful because its very high sensitivity can achieve large range for given transmission power but there are regulatory restraints on power in unlicensed spectrum.”
Today, the benefits and advantages of each technology are yet to be fully crystallised – in all markets at least some of the options are currently unavailable and in some they may never be offered because of regulatory constraints. This is creating some bewilderment among IoT pioneers, many of which are not experts or even interested in wireless technologies. Nevertheless, as the market matures, having options that closely fit requirements will be an important enabler for many IoT business cases.
“Choice promotes competition and innovation, both of which benefit the IoT industry and organisations seeking to build affordable connected products and services,” points Figueredo.
Forse sees the messages becoming clearer but different technologies continuing to exist and provide choice. “It’s impossible to simply compare the technical specifications of these competing technologies on paper and choose a winner, because requirements in the real-world change on a case-by-case basis,” he says. “There will of course be winners and losers in the race to provide IoT infrastructure, but there is also plenty of space for coexistence. In fact, a coexistence of technologies will be essential if IoT is going to be a true success.”
A note of caution is injected by Fourtet who warns that, even if all the choices and the planned developments come to fruition, not all business cases will be able to be supported. “It’s not for me to answer but there is probably already too much choice,” he says. “More than this, players aren’t constructively thinking about what they do. Some think everything is possible – it’s not. If you have a rich content application you’re limited by capacity, maybe not the cost, but there are basic Physics facts that mean you won’t be able to do everything.”
Which technology should you choose?
In the current IoT market place there are several IoT network wireless technologies to choose from. These include low power wide area networks (LPWAN), which have several proprietary iterations including Sigfox and LoRaWAN, cellular connections in the GSM mobile telephony family, now with dedicated versions such as Cat M1 for IoT and the emergent narrowband IoT (NB-IoT) specification. In addition to these further options include wireless smart ubiquitous network (Wi-SUN) technology and traditional satellite communications, Bluetooth and derivatives of Wi-Fi. We have confined our technology profiles below to the technologies that utilise radio technologies to deliver bandwidth at lower capacity than GSM cellular options.
Narrowband IoT is a new mobile technology specification developed by the 3GPP standards body to address the need in IoT for low power and infrequent data transmission devices. NB-IoT can operate in the GSM spectrum or utilise an unused resource block within an LTE carrier’s existing guard-band. NB-IoT compliant chipsets are now becoming available but they are new to market and still in the prototype stage. 3GPP has not announced plans for an NBIoT certification programme yet. Certification programmes with cellular companies have involved high fees in the past but no details are currently available. NB-IoT particularly in Europe is being seen as an important successor to 2G networks, which are starting to be retired.
LoRaWAN is a proprietary low power radio technology developed by Semtech but licensed for global use by other vendors. The technology is supported by the LoRa Alliance, which provides certification for vendor interoperability. LoRaWAN radios are commonly used in low power devices with infrequent data transmissions. LoRaWANs are typically laid out in star topology with gateways relaying messages between end-devices anda central network server.
Sigfox is a narrowband – or ultra-narrowband – technology. It uses a standard radio transmission method called binary phase-shift keying (BPSK), and it takes very narrow chunks of spectrum and changes the phase of the carrier radio wave to encode the data. This allows the receiver to only listen in a tiny slice of spectrum which mitigates the effect of noise. It requires an inexpensive endpoint radio and a more sophisticated base station to manage the network. The technology is therefore suited for connecting lowenergy objects such as electricity meters, smartwatches and washing machines, which need to be continuously on and emit only small amounts of data.
Wireless smart ubiquitous network (Wi-SUN) is a technology based on the IEEE’s 802.15.4g standard. Backed by the Wi-SUN Alliance, Wi-SUN has a third party organisation that develops tests to certify that IEEE 802.15.4g standard-based IoT equipment is both conformant and interoperable with other certified equipment. Wi-SUN networks support star and mesh topologies, as well as hybrid star/mesh deployments, but are typically laid out in a mesh topology where each node relays data for the network to provide network connectivity. Wi-SUN networks are deployed on both powered and battery-operated devices.