LPWAN Technologies for Massive IoT: The View beyond 2022

The beginning of July saw Paris welcome the LoRaWAN World Expo, the largest event to date dedicated purely to an unlicensed Low Power Wide Area Network (LPWAN) Internet of Things (IoT) network. The expo showed healthy signs: while still a small event by most metrics, it assembled all of the primary proponents of LoRaWAN technology, and reinforced again the impact of the LoRa Alliance’s work, namely the gradual building of what is LoRaWAN’s greatest strength, its ecosystem. The expo put on display a combination of complex engineering work (e.g., roaming and interoperability of networks and network types; integration with application servers), as well as commercialization for LoRaWAN’s present and future (e.g., Semtech’s LoRa Cloud Locator, which it hopes will facilitate trialing and scaling IoT and persuade a hesitant ecosystem of the benefits of its LR1110/LR1120; NNNCo’s smart label prototype based on LoRaWAN; Lacuna Space and EchoStar Mobile discussing a commercialization strategy for filling LoRaWAN connectivity gaps; and Microsoft and Amazon Web Services (AWS) in conversations on how to facilitate the job of adoption for enterprises).

The breadth of the LoRaWAN ecosystem was clearly in evidence at the expo, and the opening keynote highlighted the foundational role of LoRaWAN in connecting the IoT, as the LoRa Alliance would wish to see it, addressing up to 75% of use cases within the IoT market. However, it was also noteworthy that while the narrative of the expo itself naturally enough focused around LoRaWAN, individual discussions and vendors were equally interested in the rest of the connectivity landscape. LoRaWAN network server vendor LORIOT highlighted the initial steps in its process to integrate mioty into its existing LoRaWAN network server; and Actility was doing a similar thing on the cellular side with LTE-M and Narrowband (NB)-IoT. HT Micron was planning an NB-IoT chip as the next product on its roadmap, and tracking device manufacturer Digital Matter was planning an NB-IoT-only device in the coming months. Even much-embroiled Sigfox was equally present in participants’ thoughts, as many radio and device vendors still saw a significant opportunity in Europe and Southeast Asia in particular, especially as Sigfox-owner UnaBiz continued to talk about “LPWAN convergence” and entry into the LoRa Alliance, presenting itself not as a competitor humbled, but a partner hitherto misunderstood.

The topic of “LPWAN convergence” has grown into a buzzword because UnaBiz’s acquisition of Sigfox had stated intention in mind. There have been several ideas around what the term could actually mean from a practical perspective ever since. One theory is simply that it involves integrating Sigfox and LoRaWAN backends, which have similar—or at least overlapping—characteristics. But there seems little realistic interest in this complex engineering task from the LoRaWAN side. Another is the idea of multi-radio LPWAN devices; for instance, modules that build around both licensed and unlicensed technologies, as with GCT’s GDM7243i or the Flex Sigfox-Cellular tracker, which never took off at the time for lack of a specific business use. There is some interest around this for specific use cases in critical applications when powerful backhaul is necessary, but most of the activity here is taking place at the gateway and network level, rather than compromising Massive IoT by making it too complicated and feature-rich at the end node level.

The most convincing argument around LPWAN convergence concerns accepting connectivity technology niches, simplifying connectivity data plan sales to end customers, and simplifying how these data are fed from the network server to an application server. In particular, as the IoT becomes more pervasive, these enterprise customers are likely to use devices from large numbers of Original Equipment Manufacturers (OEMs), each requiring different transmission rates and message sizes based on their use case; the idea is to simplify how data are purchased and billed, with a neutral approach to the connectivity technology employed. Think what 1NCE did for the cellular IoT market, but with access to other network types as well. The second part of this is in simplifying how data are fed from the end node to the network server and then directly into the application server. The topic of LPWAN convergence then strongly plays into the connectivity-as-commodity argument, so that the rest of the ecosystem can think about generating value through software and services, thus removing much of the thought that an enterprise might need to give to the science experiment technology side, and allowing them to spend more time thinking about desired outcomes.

While then the LoRaWAN World Expo keynote highlighted the 75% relevance of LoRaWAN, the ecosystem participants themselves had a different focus: how to address the use case and how to integrate into application servers, regardless of the chosen connectivity technology. Indeed, the message of LPWAN connectivity technologies competing against each other has largely fallen out of fashion. Each technology is increasingly finding the niches where it is strong: ignoring one technology for another risks missing out on a very large chunk of the market, while accepting the niches will result in more value for all involved.

The LPWAN market is going through important changes. From a technology perspective, LoRaWAN, LTE-M, and NB-IoT are all highly competent in the use case areas for which they were designed. From a network perspective, each is again highly competent: cellular LPWAN rollouts have continued to grow globally, LoRaWAN has the widest range of low-cost network deployment models, and both the cellular and LoRaWAN sides are making significant strides with Non-Terrestrial Networks (NTNs) as exemplified by recent announcements from Telefónica, Soracom, Senet, UnaBiz, EchoStar Mobile, and many others. Other unlicensed technologies, such as Sensus, mioty, Sigfox, Dash7, and many others help fill specific gaps where no other technologies will provide the same level of value. New standards, such as DECT-2020, will do a similar thing while simultaneously bridging the gap between the licensed and unlicensed connectivity markets.

The conversation around LPWAN convergence is an open one and likely to create many headaches. However, discussions in this tenor highlight a more mature conversation happening at an ecosystem level: the IoT is not a question of one technology or another, but of a blended approach with the goal of generating data that can be used for a particular outcome. The result is that features of connectivity technologies become simultaneously more important and less important: they become more important to the IoT ecosystem solution providers building and assembling a solution that can generate the highest level of value, while becoming less important for those adopting the technology whose focus will continue to shift further toward the software side of the market. What this is likely to mean is a realignment—further in favor of solution integrators—of how IoT is sold to enterprises, as all components of a solution are bundled together before being sold to a customer. It will also mean a greater focus on managed services to keep the technology systems up and running, while enterprises continue on with their daily operations. These are areas for the IoT ecosystem to be aware of.