There is a lot of buzz in the media recently about contact tracing apps for smartphones. These apps are touted as being a new solution for pandemic management in the digital age. By utilizing smartphone Bluetooth capabilities, contact tracing apps can keep tabs of people’s contacts and create an easily searchable database in case a person is identified as infected.
However, contact tracing apps have a few major shortcomings. First and foremost, contact tracing apps rely on there being a high smartphone penetration rate in the population. In developed nations and high-income areas, smartphone usage is basically ubiquitous. In developing nations, smartphone usage is much more sparse. Even within developed nations, smartphone ownership is stratified based on race, gender, and income. This means that smartphone-based apps for contact tracing cannot be a universal solution.
Further, smartphone-based apps rely on WiFi or cellular networks to communicate data to servers. There are many regions in both developed and developing countries that do not have stable WiFi or cellular service. Smartphones also eat up a large amount of power between keeping location services on and continuously transmitting data.
LoRaWANS as a Geolocation Solution
Long-range wide-area networks (LoRaWANs) are a kind of low-power wide-area network (LPWAN) that is specifically designed to maximize transmission distance while minimizing power consumption LoRaWANs are very efficient for geolocation services because they do not rely on applications the require a high amount of power such as GPS or WiFi.
First, a quick recap of LoRaWANs. LoRaWANs actually have two parts. LoRa is the name given to the lower-level physical technology that can modulate continuous radio signals into discrete packets of bits. LoRaWAN refers to the higher-level abstract protocols that govern how LoRa signals interact with gateways, encryption, etc.
LoRaWANs work by each ‘node’ sending periodic signals to a gateway. Gateways can listen for specific nodes and can be configured to communicate downstream with nodes. This kind of ‘star topology’ is what allows LoRaWANs to communicate over long distances with minimal power consumption.
So how does geolocation with LoRaWANs work? Through a process known as Time Difference of Arrival (TDOA), also known by the more familiar name ‘triangulation.’ (technically, LoRaWANs use ‘multilateration’ instead of ‘triangulation’, but the basic principle behind the two is similar).
Here is how LoRaWAN multilateration works. A LoRa end device can be located if at least three gateways pick up its signal. Gateways can calculate how far away the device is by either measuring how much the signal has dispersed (signals disperse over time across distances) or by measuring the time it takes for the signal to get there. The latter method is normally more reliable as the speed of light (information transfer) is continuous in a given medium, such as the atmosphere.
If the physical locations of the three gateways are known, then finding the LoRa-end device is as simple as drawing a circle around each gateway with a radius corresponding to the measured distance. The LoRa-end device can then be located in the small portion of space where the three circles overlap. Technically, most multilateration services construct hyperbolas instead of circles and then determine the point of intersection of the 3 curves. However, the general method is the same.
So, using this multilateration method, LoRaWANs can pinpoint the location of objects as long as the node can be detected by at least 3 gateways. The natural conclusion of this setup is that LoRaWAN geolocation networks are more accurate if there is a denser distribution of gateways. In short, the more gateways, the more chance that three of them will pick up a node’s signal, and the more accurately it can be located.
How Accurate Is LoRaWAN Geolocation?
LoRaWAN accuracy can range anywhere between 20-200 meters. The exact accuracy depends on a few factors such as how dense the LoRa network is and how large the physical distances are. Due to propagation errors in signals, distance and accuracy have an inherent tradeoff. Having a denser LoRa network can balance out this trade-off and keep accuracy high. It is also important that the physical locations of the gateways are accurately measured with respect to each other. Otherwise, there can be errors that offset the curve intersection.
LoRaWAN accuracy is also affected by the surrounding physical environment. In large, open, flat areas, LoRaWAN geolocation tends to be more accurate as there is less interference from geography or other structures. In urban environments, geographic variation or large structures can affect accuracy. Thus, it follows that urban environments would require a denser LoRa device distribution to get comparatively accurate results than the same size of rural land.
According to research, in clear multipath free (interference-free) environments, LoRaWAN geolocation can be as accurate as 3m. The story is different in multipath environments. In multipath environments such as a dense urban locale, accuracy can range between 60m-80m, provided that signal interference and timestamp errors are accounted for. This is still in a suitable range for figuring out general locations and directions of things. Research has also indicated that LoRaWAN multilateration can track a moving vehicle in real-time to a precision of 171m.
CareBand’s SafeTrack wearable public health solutions can use LoRaWANs in two ways: only focused on context or focused on context and geolocation. Context-based applications include something simple like a geofence for proximity monitoring. Simple contextual clues can be used to get a rough idea of where SafeTrack users are. However, if the LoRa network is dense enough, SafeTrack devices can use geolocation services to sharpen up the picture and more accurately identify position.
LoRaWANs for Contact Tracing
LoRaWANs are an ideal tool for contact tracing networks because they are scalable, have a high degree of accuracy, and also have no external hardware requirements aside from the basic LoRa end-devices and gateways. The ability to both track location using context and geolocation services is useful for many applications, from proximity monitoring, geofencing, contact tracing, and locating individual device wearers. The minimal power requirements and low-cost make LoRaWAN-based contact tracing solutions highly efficient and cost-effective.
