Last month saw the publication of an in-depth report from experts at Real Wireless comparing the relative merits of different radio technologies for Low Power Wide Area (LPWA) applications. The report casts doubt on the suitability of LoRa for mass adoption in unlicensed spectrum.
Real Wireless LPWA report: click image to download
Our take: LPWA technologies using spread spectrum techniques, like LoRa, will struggle to reach commercial device volumes because of interference issues. And these networks may also act as "bad neighbours", actively polluting established applications in unlicensed Industrial Scientific and Medical (ISM) spectrum. By contrast, the report finds that Ultra Narrow Band (UNB) systems such as Sigfox and Telensa do not have these problems.
We have been surprised by the popularity of LoRa with mobile operators given these problems, which are well known to industry insiders. However, to date, all of the deployments of this type of technology have been at such a small scale that users will not necessarily have seen the problems.
But spread spectrum works fine – it’s used in 3G.
Yes, but 3G took a huge effort to get it working. When it did finally work, battery life was poor and devices were very expensive. This wasn’t the normal growing pains of a new technology but a reflection on the inherent complexity of 3G technology.
Underlying spread spectrum technology is a concept called “orthogonality”, which is a measure of how communications from different users can be distinguished from each other. In order to achieve this, 3G assigns complex codes to each user. These codes are carefully managed as any communications that don’t use the correct codes cause interference. This works fine in spectrum that is owned by a single operator but isn’t workable in shared spectrum such as that used by LoRa.
In addition, 3G spends a large part of its resources fine tuning the power levels and frequency of each device continuously to maintain orthogonality. This uses up battery life and adds to device cost which goes against the goals of LPWA.
So why are some network operators adopting LoRa when they should know better?
Many operators feel under pressure to show that they have a workable strategy for IoT especially with the rise of new players such as Sigfox. However, the technology that will enable them to support IOT alongside their existing infrastructure is some time away. As a result, operators have had to adopt a defensive strategy and have simply grasped at the easiest technology to buy off the shelf in the full knowledge that they won’t have to roll out full scale services until the new technology (called NB-IOT) comes along.
Now, more cynical people than me may suggest that it is in the operator’s interest to promote a technology that they know has limited scope to compete with them longer term. That would never happen would it?