Where to next for WAN IoT connectivity?

Three principal hardware features characterize the enterprise internet of things in comparison with traditional connected consumer devices: low power, small device size, and small data packets. To meet these unique requirements, the IoT required a new set of connectivity technologies that did not focus on uninterrupted connections to the network and higher bandwidth for voice, video, and browsing. Enter low-power wide-area networks (LPWANs), starting with the first wave of Sigfox and Cycleo (the company behind the LoRa modulation IP) around 2010, using unlicensed spectrum and continuing with the 3GPP’s standardization of NB-IoT and LTE-M on licensed spectrum in 2016.

While LPWANs have seen enormous growth in the past few years, they face challenges.

First, the number of connectivity technologies available for the IoT has created a quagmire as technology providers and enterprises look to understand the relative benefits and drawbacks of LPWANs, as well as to assess claimed attributes in relation to real-life performance. The quagmire is responsible for many abandoned proofs of concept, the lack of confidence in IoT capabilities, and, crucially, for the slower-than-anticipated adoption.

Second, LPWANs remain in a state of flux: None have settled into an easily definable role, and experimentation and technology enhancements in the form of new standards and guidelines are ongoing.

This article provides an overview of the LPWAN market as it currently stands: the technologies, the applications, and the ecosystem dynamics, which, at the end of the day, underpin the success of the network.

Evolution of IoT connectivity. Click for a larger image. (Source: ABI Research)

Unlicensed LPWAN

Sigfox and LoRaWAN are the most well-known and adopted technologies in this segment, though others exist. These technologies operate on sub-gigahertz unlicensed ISM spectrum, which allows them to achieve long-range coverage and high penetration capabilities. The unlicensed nature of the spectrum in theory means that connectivity prices can be kept low — critical for return-on-investment calculations.

On the downside, ISM spectrum is subject to more duty-cycle restrictions, lower service levels (such as guaranteed message receipt), and lower security than encrypted licensed networks. Unlicensed LPWANs are considered to have had a first-mover advantage due to their entry into IoT markets before cellular networks, but they are increasingly having to justify their role and niche in IoT.

Sigfox
Sigfox has the longest range, highest penetration capability, and lowest power consumption compared with other mainstream LPWAN technologies. Its purpose was always to enable an ecosystem to build the smallest and lowest-cost connected devices, transmitting extremely small data packets — up to 12 bytes on the uplink — mostly in a single direction, from sensor to cloud.

The company also differentiates itself with its business model: It sells network hardware and a cloud to nationally established Sigfox operators, who themselves sell connectivity to end customers. By providing the network hardware for its operators, Sigfox retains control over network parameters globally. This enables another two principal advantages — namely, its ability to create a single global network for roaming and, consequently, a very simple hardware design. Its strengths have seen it win high-profile contracts in asset tracking, metering, and security alarms, among others.

Still, few were flabbergasted by the news of Sigfox’s receivership in January 2022. It had filed losses in 2020, its South African network entered rough waters in 2021, and very few were convinced by its increasingly low-price predictions for its devices: $1 buttons and — one week before news of its receivership broke — $0.30 tracers with a 13-year battery life seemed more marketing than reality. It also struggled to justify its €10/year connectivity prices for, fundamentally, very small quantities of data and a lower service level compared with licensed networks, while other LPWANs’ pricing decreased significantly. Its future, at time of writing, is uncertain.

LoRaWAN
Promoted by Semtech (the owner of the protocol IP) and the LoRa Alliance (ecosystem and technology promoter), LoRaWAN targets the development of simple low-cost devices with a higher packet size and throughput rate than Sigfox (243 bytes and 50 kbits/s) and with fully bidirectional communication and unlimited messages. The number of devices connected on the network has grown exponentially in the past few years, driven by an increasingly healthy ecosystem thanks to strong market education.

The ecosystem has two further strengths. First, Semtech gives the people what they want. Two examples of this are the LR1110 chip in 2020, which proved a great success, and Semtech’s additional support in the same year for long-range frequency-hopping spread-spectrum (LR-FHSS) — described to me recently as “Sigfox on steroids” — supporting a broader range of transmission types, higher penetration, higher device density, and non-terrestrial networks (NTNs, aka satellites). Second, Semtech is actively engaged with new initiatives. Recent examples of note include people-powered fast-growing Helium, Amazon Sidewalk, and extensive engagements with low-Earth-orbit (LEO) satellite companies.

It hasn’t all been smooth sailing for LoRaWAN as it looks for its safe harbor, however. As cellular LPWANs have come in vogue, its popularity could wane — as shown by Bouygues Telecom announcing the shutdown of its Objenious LoRaWAN network at the end of February 2022 in favor of NB-IoT and LTE-M. LoRaWAN appears to be pivoting more toward private network deployments, where LoRaWAN infrastructure provider Kerlink currently has 80% of its deployments. Arguably, this is where LoRaWAN can differentiate itself from its LPWAN competition.

Licensed cellular and LPWAN

In response to unlicensed LPWAN networks, the 3GPP standardized the NB-IoT and LTE-M to fight for the licensed network corner. Licensed networks tend to be more expensive to run, as telecoms want to recoup the cost of investment from buying national spectrum, but are also more secure and come with higher service levels. Initially, the 3GPP was a little late to the LPWAN party and was playing catch-up to unlicensed networks, while telcos worked out their deployment and monetization strategies and sought to build an ecosystem from scratch. Cellular networks for IoT have faced many challenges: Roaming agreement challenges, pricing models, the ability to support small deployments, and understanding of technology capabilities are among only a few. However, these are getting ironed out as the ecosystem matures, accelerating adoption.

LTE-M/NB-IoT
These technologies permit small data payloads and a higher data-throughput speed compared with unlicensed networks. NB-IoT tends to be deployed primarily for stationary or little-moving sensors requiring very small data transmission. Due to its lack of support for cell-tower handoffs, it is usually considered less suitable for mobile applications like real-time asset tracking. Both NB-IoT and LTE-M support extended discontinuous reception (eDRX) and power-saving mode (PSM), making these very low-power–consumption technologies.

Customers are generally attracted by the high service levels that come with cellular networks. These include message receipt guarantee, interference protection, network security, bidirectional communication, widespread coverage once telecoms fulfill their deployment strategies, a strong hardware developer community, a strong geolocation developer community, and “future-proofness,” as both LTE-M and NB-IoT have been written into 5G standards.

A current prominent use case is asset tracking on LTE-M, but the technologies are popular for any type of low-power connected sensor device. As pricing and roaming are figured out, these technologies have become the most high-growth opportunity in IoT.

2G/3G
These “legacy” networks remain remarkably useful for IoT. It is not uncommon to see modules offering LTE-M/NB-IoT with fallback to 2G/3G when high mobility is needed. 2G/3G has the benefit of global coverage, a status that NB-IoT and LTE-M do not yet have due to the high cost of deploying new infrastructure globally, which requires a more selective and phased approach. 2G/3G networks are approaching sunset, however (though not for years, in some cases); a popular alternative may be Cat-1 due to its ability to leverage existing 4G/LTE infrastructure.

Follow closely

The technologies described above can now be considered mainstream within the IoT. However, that does not mean that we are at the end of the evolution of networks and connectivity for the IoT. Two further items are noteworthy.

Mioty
Licensed by BehrTech, unlicensed network Mioty has already won several contracts against its mainstream competition, with technical capabilities far exceeding any other LPWAN on the market. Range, penetration, ability to support packet loss, sensor mobility, sensor density — Mioty is ahead in all of these. While it currently lacks the global coverage and developer ecosystem to support it and become a long-term commercial option, Mioty’s trajectory is worth watching.

Satellite
The new connectivity frontier is space. Numerous companies have entered the LEO market for connecting remote devices, and in the second half of 2021, there was a scaling up of satellite launches to show for it. Many different approaches exist, such as LoRaWAN or NB-IoT to satellite via gateway, or direct low-power sensor-to-satellite. One initiative to watch comes with 3GPP Release 17 in Q1 2022, which standardizes the extension of NB-IoT and LTE-M for use in NTNs. There are barriers in the way of satellite use for low-power connectivity, but in the medium term, the approach should enable greater opportunities for the IoT industry.

Fundamentally, suitable connectivity choices vary based on use-case requirements. “Critical” or “massive,” remote or urban, stationary or mobile — these are only a few parameters that will guide choice. Seek help in finding what works best for your situation.

As part of the Enabling Platforms research team at ABI Research, research analyst Tancred Taylor focuses on asset tracking and visibility, which includes data, trend, and forecast analysis of the devices; platforms, networks, and services empowering the emerging technology market; and the business transformation that IoT tracking will enable. Tancred’s coverage spans many industries, from construction and heavy equipment to supply chain and packages and parcels.