The third pillar of 5G, mMTC (massive machine-type communications), will not be delivered until 3GPP Release 17 comes along at the end of 2021 (see separate item), and even then it is not clear it will meet the ambitious density promise of one million devices per square kilometer. For this reason, consideration has been given to including LoRAWAN within the 5G framework, even as an option within licensed bands.
Although LoRAWAN has been considered a competitor to mMTC, it has hitherto operated only in unlicensed bands and its attraction is that it has gained critical mass there, while recently being enhanced with functions that would make it more acceptable to the 5G community. The 5G standards so far promise to address ultra-low latency and high data rate use cases, but when it comes to massive IoT (the third corner of the triangle), the broader 5G movement is concerned with the whole future of wireless communications, both indoors and outdoors.
The advantage LoRAWAN has is that it was designed specifically for the outdoor IoT as a low power WAN technology, for applications requiring large geographical range up to 10 kilometers or more, but where total data volumes and individual bit rates are quite low.
By addressing this use case, the underlying LoRa radio developed by Semtech was made low cost both to deploy and operate. The radio technology was available in low specification silicon, cheap enough for inclusion even in simple environmental sensors, while being so light on power that batteries last for years, governed more by shelf life.
Initially, the underlying LoRa radio system was incorporated by various vendors in their own systems with proprietary transport of data between locations, enabling full working products to come to market. Then the LoRaWAN Alliance was formed to standardize the whole protocol stack so that any LoRa sensor could communicate with any LoRaWAN gateway, enabling users to adopt a best-of-breed approach for devices if they wanted.
This full LoRaWAN layer was available a good half decade before the 5G ecosystem and this window has enabled a great variety of sensors and end-to-end packages to be deployed, amounting to around 650,000 gateways and 180m devices globally by 2020. It has also been deployed by over 100 MNOs or telcos, with strong advocates including Orange, Swisscom, Bouygues Telecom, Comcast, Proximus and Deutsche Telekom.
Many of these operators are deploying LoRAWAN alongside 5G options for LPWAN, which are currently inherited from LTE (that is NB-IoT and LTE-M). Such hybrid strategies are encouraged and welcomed by the LoRAWAN Alliance, which agrees that the two meet different use cases and that its stack is not the best fit for all of them.
Although there are variations between operators and regions, LoRaWAN tends to be preferred where reach and robustness are paramount but high throughput and large numbers of data transmissions are not required. This includes some smart city or smart building projects, as well as water meter connectivity.
On the other hand, electricity smart grids generate larger amounts of data and in more frequent transmissions, which the LoRaWAN Alliance admits is better suited to cellular IoT with the focus on large-scale machine connectivity.
However, LoRaWAN has gained traction for a variety of use cases for unlicensed spectrum, both among cable companies and fixed operators, because it is the lowest cost option that meets the requirements.
Then LoRaWAN will continue to expand in its existing stronghold of smart agriculture, for applications such as soil moisture or pH monitoring to improve crop yield, or assessing livestock condition to boost yields of milk or meat. LoRaWAN also scores in environmental monitoring, where again it enables development of affordable WANs in areas where cellular connectivity is not always available.
This also includes those IIoT applications where high throughput and low latency are not required, around process or system monitoring rather than real time control and automation. The technology appealed there partly because of the tolerance to physical barriers such as concrete walls at LoRaWAN frequencies, 863–870 MHz in Europe and 902–928 MHz in the US for example.
It is true some of these use cases in areas served by cellular could be threatened by 5G, except for the fact that the latter may need LoRaWAN to fill in. Recent studies have looked at the extent to which LoRaWAN can provide some of that promised density and found that it can readily contribute around 10%, at least for the uplink.
That may sound disappointing but will probably ensure that LoRaWAN remains part of the mix for the medium future, while retaining its hegemony in its agricultural heartlands, especially in the vast expanses in countries such as Australia, the US and Russia, where provision of ubiquitous cellular connectivity would be uneconomical.