|
Title
|
|
|
|
Localization with low power wide area networks
| |
|
Author
|
|
|
|
| |
|
Abstract
|
|
|
|
| We can hardly imagine a world without localization these days. Travelling to a place you have never been, finding the best restaurants nearby, figuring out the current location of that thing you ordered from China three weeks ago, it is all just a few clicks away on our smartphones or computers. Often, we get this location information through a Global Navigation Satellite System (GNSS) such as Global Positioning System (GPS) or Galileo. These systems are ubiquitously present in devices we use in our every day lives, e.g., our smartphones, smartwatches and cars. The demand for location information is also rapidly increasing in the so-called Internet of Things (IoT). IoT aims to interconnect a wide variety of objects, ranging from temperature sensors on mobile cooling containers to garbage bins in a city. Low Power Wide Area Networks (LPWANs) play a very important role in communicating the measurements from these objects to an end-user, as they are frequently responsible for the wireless link between them. Of course, a user often needs to know where their objects are located in order to interpret the measurements they have received. Hence the need for localization solutions in IoT. An important constraint on the communication and localization solutions is that they must be as energy-efficient as possible, because IoT devices generally operate for multiple years using small batteries. This, and the fact that GNSS can normally only be used in outdoor environments, raises the question if adding GNSS hardware to an IoT device is the best solution to obtain location information for that device. An alternative localization solution can be offered by the wireless LPWAN communication link that is required on an IoT device anyway. Signal characteristics such as the Received Signal Strength (RSS), timing measurements or phase measurements enable the use of GNSS-less localization methods. Nevertheless, research is required to gain insights on the achievable location accuracy of these methods when applied to an LPWAN. Only then will developers be able to make an informed choice between GNSS and GNSS-less localization for their own IoT applications. This thesis aims to gain these insights using real LPWAN measurements rather than simulations and theoretical analysis. Accordingly, I started my research by collecting measurements in LPWAN deployments in large urban and rural areas. The resulting datasets have proven their worth not only for my own work, but also for that of other researchers. In the past four years, this has enabled both LPWAN localization and communication to be examined from many different perspectives. I continued my research by analyzing what location accuracy can be achieved in LPWAN deployments that are not optimized for localization purposes. Specifically, I evaluated multiple RSS-based localization methods in a large urban area. Next, I evaluated if a significantly lower estimation error could be achieved if the LPWAN deployment in that urban area were to be optimized for localization. To that end, I installed my own LPWAN infrastructure and experimented with probabilistic combinations of Time Difference of Arrival (TDoA) and Angle of Arrival (AoA) localization. With these experiments, I gained realistic insights into the achievable accuracy of LPWAN localization. The last chapters in this thesis consider those insights to define the application potential of LPWAN localization in the IoT landscape. I analyze how a developer can choose between GNSS and GNSS-less localization considering the balance between technological constraints and application requirements such as location accuracy, update rate and battery lifetime. Furthermore, I describe how LPWAN localization can be applied in a multimodal setting, where different technologies and methods are applied to meet the dynamic application requirements of IoT devices moving between different environments. |
| |
|
Language
|
|
|
|
English
| |
|
Publication
|
|
|
|
Antwerp
:
University of Antwerp, Faculty of Applied Engineering
,
2022
| |
|
ISBN
|
|
|
|
978-90-5728-730-5
| |
|
Volume/pages
|
|
|
|
xx, 134 p.
| |
|
Note
|
|
|
|
:
Weyn, Maarten [Supervisor]
:
Berkvens, Rafael [Supervisor]
| |
|
Full text (open access)
|
|
|
|
| |
|