Your browser is not supported. Please update it.

26 September 2019

LoRa Alliance and WBA unveil LoRa-WiFi integration paper, miss killer-apps

The Wireless Broadband Alliance (WBA) and the LoRa Alliance have published a whitepaper that outlines how best to deploy LoRa alongside WiFi. The partnership is interesting, and another admission that LoRaWAN has found a niche in the same world that WiFi inhabits, that is, in campus-type deployments, to such an extent that a whitepaper explaining how best to integrate the two was compiled by two standards groups.

From the perspective of a business function, the technology that has carried the bits and bytes to the intended location matters very little. The C-Level suit looking at their shiny cloud-based business dashboard will consider WiFi and LoRaWAN to be one and the same, so long as they both correctly carry the data that power these IoT functions. However, from the networking perspective, such wrinkles are cavernous, and worthy of proper exploration.

This perhaps leads to a philosophical discussion of what actually constitutes a wide-area network. Given that our definition of LPWAN includes a mesh technology like Wi-SUN, you can probably guess where this is going. Add to this that the upcoming plans for 5G networks comprised of small cells and radio frequencies that can’t penetrate walls, and so are very comparable to WiFi, and it seems reasonable to conclude that if you strung enough WiFi networks together, you could easily view the joined-up whole as a WAN.

So, where do you draw the line between a massive city-wide or campus-wide network being a WAN? Further, if it is aimed at IoT applications, has it strayed into LPWAN territory? Would low-power WiFi then be considered an LPWAN technology in those sorts of environments?

Perhaps that rabbit-hole is better suited for a few pints at a networking event, but you can see how blurry things get when we start combining technologies. In a few years, does LoRa simply become a low-power extension of WiFi that a networking gateway handles, exposing it to applications as if it were simply another IP-enabled device? Existentially, does that matter?

There are plenty of companies out there making use of LoRa’s LPWAN functionality, connecting devices that are multiple kilometers from a gateway, but data on the proportion of short-range or gateway-dense applications is very hard to come by. It would be very interesting to know the number of LoRa devices that could actually be served by a sufficiently low-powered version of WiFi or Bluetooth, in terms of their proximity to an access-point.

“The reality is that no one single technology is going to fit the billions of IoT use cases,” said Donna Moore, CEO and Chairwoman of the LoRa Alliance. “It is collaborative initiatives like this one with Wi-Fi that will drive innovation to solve important issues, leverage an even broader range of applications and, ultimately, ensure the success of global mass IoT deployments in the future.”

“Wi-Fi and LoRaWAN are two important technologies utilizing the unlicensed spectrum, and they already address a large proportion of IoT use cases,” said Tiago Rodrigues, General Manager, WBA. “The Deployment Synergies paper highlights the ways in which these technologies are impacting private-public business models and enabling IoT services, while also identifying ways in which the technologies complement one another and can be used to further expand the Internet of Things.”

The paper itself, Wi-Fi & LoRaWAN Deployment Synergies: Expanding Addressable Use Cases For The Internet of Things, was built using input from MNOs, telco equipment manufacturers, and “advocates of both connectivity technologies.” Collectively, the pair say that it illustrates new business opportunities that are created when you pair the two together.

One of the main thrusts is ‘Massive IoT,’ which the document says “are less latency sensitive and have relatively low throughput requirements, but they require a huge volume of low-cost, low-energy consumption devices on a network with excellent coverage.” While WiFi covers short and medium range use cases at high data rates, LoRaWAN is for long range and low throughput, especially “in hard to reach locations, such as temperature sensors in a manufacturing setting or vibration sensors in concrete.”

To this end, the three key use cases that the paper highlights are Smart Building / Smart Hospitality, Residential Connectivity, and Automotive and Smart Transportation. We have a bone to pick with the claim that “both technologies have been deployed for decades throughout buildings,” though.

In the first, the pair make the valid point that LoRaWAN can be used for smoke detection, asset and vehicle tracking, and room occupancy sensing, and that WiFi is used here for security cameras and high-speed internet access. However, there’s not really a killer-application for LoRaWAN here, as a building is perhaps the best place you could possibly deploy a dense WiFi network. Similarly, you could use Bluetooth instead of LoRaWAN in many of the examples, or a wired power and data connection to a static device, and so the main use of LoRaWAN here ends up being possibly tracking assets. This isn’t a super compelling argument.

Similar problems arise in Residential Connectivity, as again, we’ve not really delineated WiFi and Bluetooth from LoRaWAN in either power constraints or long ranges. Further, the argument that you could use LoRaWAN in the home for leak detection and fuel monitoring rubs up against the actual decade of experience (and lack of enthusiasm) that Zigbee and Z-Wave have in that sector, and the concept of linking LoRaWAN picocells into “neighborhood IoT networks” just straight up slams into the whole business case problem. Exactly how does this get monetized in a realistic time frame?

The final transportation argument also butts into the killer-app problem, in that “fleet tracking and vehicle maintenance” wasn’t in Riot’s use-case based LPWAN forecast for a reason. Telematics data is not something that you want to be backhauling using LoRaWAN, and if tracking is the main requirement, most new fleet vehicles sold these days come with fully-fledged cellular connections. Sure, third-party boxes can make use of LoRaWAN there, but the automakers are slowly getting their act together here, and will leap at the opportunity to provide such services and cut out the middleman.

To be clear, it’s not a bad paper. It’s worth a quick read, but if the goal was to prove that LoRaWAN was an indispensable connectivity technology, it somewhat missed the mark. There’s plenty of market to go around, and LoRa-based solutions are certainly going to have some success, but the LPWAN community needs to focus on environments where LPWAN technologies are the only suitable solution to the problem, and nail those before expanding into more crowded territory.

The list of companies involved in the paper reads as follows: BT, Boingo, BSG Wireless, Charter Communications, Connexin, Eleven-X, ER-Telecom, Orange, Tata Communications, Unity Media, Objenious, Semtech, Syniverse, Abeeway, Actility, BSG, Kerlink, Maxima Telecom, Microshare, Orbiwise, Senet, Siradel, Skyhook Multi-Tech, the Centre for Development of Telematics and Digital Catapult.