LPWA-Disruptive new networks for IoT

Low power wide area networks are one of the most far-reaching trends in M2M and IoT, writes Aapo Markannen.
Published: Tue 19 Jul 2016

Machina Research has for a long time identified the emergence of Low Power Wide Area (LPWA) networks as one of the most far-reaching trends in M2M and IoT, and in the course of 2015 these technologies have truly come of age. At the moment, practically every technology supplier, service provider, and enterprise that can be regarded as the cutting edge of IoT is trying to understand how LPWA will enable, or impact, its business.

The technologies that enable LPWA networks are far from identical. These technologies can be categorised under six main groups: LoRa, UNB, random phase multiple access (RPMA), cellular, weightless, and solution-centric LPWA.

These networks meet both of the following criteria:

Low Power: The technology is capable of delivering multiple years of device operation on a single AA battery, assuming hourly application readings and factoring in the effects of battery self-discharge and degradation.

Wide Area: The technology is capable of delivering at least 500 meters of signal range from the gateway device directly to the endpoint, assuming challenging deployment conditions – such as urban and underground environments.

Apart from these criteria, there are other characteristics that can be seen as typical of LPWA. These include advantages that are of a secondary (increased network capacity) or non-technical (reduced costs) nature, as well as technical trade-offs that are a consequence of being able to meet the two defining criteria:

  • High Endpoint Density

  • Reduced Hardware Costs

  • Reduced Connectivity Costs

  • Low Data Rate

  • Constrained Latency

  • Limited Mobility

Importantly, LPWA is technologyagnostic, meaning that it applies both to new networking technologies that have been specifically designed to deliver LPWA-style performance and to incumbent ones that are adopted to do so, on an iterative basis.

Several market issues remain open

A number of market issues will define the LPWA industry’s future direction, including the relationship between LPWA technologies and radio spectrum; the commercial models that are used to the deploy LPWA networks; the role of battery life as an application requirement; as well as the role of downlink communication as an application requirement.

Spectrum use: LPWA networks can be rolled out in both licensed and unlicensed radio spectrum, depending on technology. Each approach comes with its own pros and cons:

LICENSED SPECTRUM FOR LPWA

Pros

• Freedom of network usage

• Potential to use existing cell sites

• Fast start-to-blanket coverage

Cons

• Reliance on spectrum holders

• Uncertain service propositions

• Spectrum overheads

UNLICENSED SPECTRUM FOR LPWA

Pros

• Fast time-to-market

• Enablement of new providers

• Viable BYO option

Cons

• Limited security

• Lack of multinational players

• QoS and capacity issues

 

Deployment model: There is no uniform model for deploying LPWA. Different models can be broken down along two dimensions, according to who is permitted to access the network and what level of territory is covered.

Battery life: Arguably the most opaque parameter related to LPWA technologies is battery life. On this front, as in most other technical aspects, it is advised that all enterprises test and trial LPWA more carefully than their other connectivity options. In addition, enterprises should consider proactively employing a “minimum viable lifespan” approach to their applications, instead of trying to squeeze as much battery life out of the devices as possible.

Downlink capability: Another very significant parameter in the LPWA market is downlink capability, as the range of IoT applications that require only uplink communications from the endpoint to the gateway is likely to be limited. Besides the technical characteristics of the networks, the use of spectrum plays a large role in determining downlink flexibility. All in all, downlink capabilities can be expected to become a key selling point for LPWA alternatives, as well as a premium feature for the network providers that are able to offer it to their customers.

LPWA for utilities - transformative potential

For utilities, LPWA has undeniable potential to drive technological transformation. This is especially the case in water, gas and municipal heating, which are all areas where smart metering remains a remarkably nascent concept, due to the lack of adequate connectivity options. The worldwide availability of new networks that are able to cost effectively serve large numbers of battery-operated meters in underground or otherwise difficult locations will be a definite business enabler for firms operating in the said sectors. In addition to connecting metering devices, LPWA can bring further operational gains by allowing water, gas and heating utilities to employ sensor-based condition monitoring throughout their distribution infrastructure. In these markets we are talking about deployments that would not be feasible in the first place without LPWA.

IoT and electricity infrastructure

In electricity, the picture is more complex. Electricity meters do not need to rely on batteries, and the economics for getting them connected to begin with are clearer than in water for example, so there are already a variety of connectivity solutions enabling smart metering. The likes of powerline, ZigBee, as well as the traditional cellular (2G/3G/4G) networks have all been used widely in metering rollouts to date, albeit with somewhat mixed success in terms of technical robustness and cost effectiveness. In the meantime, at grid level, electricity utilities need to address a different set of pain points. While the objective in metering, in principle, is to maximise the number of connected endpoints and minimise the overheads of doing so, the issues in grid infrastructure have to do more with coping with the vast volumes of data that is being generated in substations and other distributed “islands” of big electricity data, often beyond the reach of the fibre-optic backhaul. This is an area where the evolution towards 5G holds a lot of promise, but overall, all the real game-changers can be found outside of the connectivity layer. Smart grid is one of the most compelling verticals for edge intelligence and fog computing – two intertwined concepts that refer to the processing and filtering of data at or close to its source, instead of transmitting all the way to the enterprise level for backend analytics.

Towards virtual power plants – connecting homes and buildings

A virtual power plant (VPP) – controlled over abstracted software tools and systems – can provide much more targeted and sophisticated load-balancing and demand response measures across the electricity networks. Besides the “smarts” added to the meters, grid elements, and the enterprise backend, the end-devices that consume the supplied energy are a critical part of the VPP concept. In this context, the utility sector’s fortunes will largely depend on how the presently confusing landscape for in-building connectivity will ultimately play out. The recently released Thread protocol warrants extra attention in both residential and business settings, standing a good chance to muscle ZigBee and Z-Wave out of the market. Meanwhile, the upcoming iterations of Bluetooth (mesh networking) and Wi-Fi (sub-GHz spectrum) may also prove major enablers for in-building connectivity and thereby energy management. Furthermore, this is another domain where LPWA is expected to find significant adoption. For example, if an appliance manufacturer – in a close partnership with electricity utilities – wants to implement a solid demand-response capability for every device of its product model then it will essentially have to invest in a wireless WAN technology, because there will be no single short-range option it could rely on for universal connectivity. All in all, the outlook on IoT connectivity varies according to which part of the value chain one decides to look at. LPWA will never be able to address all connectivity needs that a utility company may have, but that does not change the fact that as a whole these networks are a very exciting addition to the technology toolkit.

 

Aapo Markkanen is a principal analyst at Machina Research.

This article first appeared in Metering & Smart Energy International Issue 6 2015.

 

Related Webinar