Doctoral Thesis
Localization, Tracking, and Data Transmission Using LoRa
Doctor of Philosophy (PhD), Murdoch University
2024
Abstract
Current remote Internet of Things (IoT) monitoring systems generally use GPS, WiFi, and mobile cellular network technologies. However, these technologies face challenges related to range, battery life, and cost-effectiveness in remote areas. In this context, LoRa (Long Range) is a promising technology due to its long-range, extended battery life, extensive coverage area, and ease of implementation.
This thesis investigates the use of LoRa for localization, tracking and efficient data transmission. The literature review addresses challenges like low accuracy, offline training, implementation complexity, and limited data rates in LoRa-based localization, tracking and data transmission systems.
The first contribution is a range-based localization technique using machine learning models that incorporate features such as received signal strength indicator (RSSI), spreading factor, and signal-to-noise ratio. This technique is enhanced by modified trilateration for more precise target node localization, achieving improved accuracy using multiple signal features.
A method for localization using a single mobile LoRa gateway is explored. Employing particle filtering and machine learning, this approach maps the distance between a target node and the gateway based on RSSI, enabling active searching and passive monitoring without the need for multiple gateways.
Unsupervised techniques are used for localization and tracking with symbolized LoRa signal features. Maximum entropy partitioning and D-Markov machines extract temporal patterns combined with adaptive trilateration for localization in dynamic environments. This method achieves simultaneous learning and estimation for localization and tracking.
Finally, a novel pipeline for LoRa-based data transmission is proposed, emphasizing data minimization and quality preservation. This pipeline employs a Siamese network for similarity indexing and the HARQ transmission protocol, achieving low payload and downlink.
Experiments were performed in semi-line-of-sight settings to validate the efficacy of the proposed techniques, showcasing their applicability in low power wide area network connectivity. Such systems have utility in wildlife monitoring and surveillance, cattle and equipment tracking.
Details
- Title
- Localization, Tracking, and Data Transmission Using LoRa
- Authors/Creators
- Khondoker Z Islam
- Contributors
- Ferdous Sohel (Supervisor) - Murdoch University, Centre for Crop and Food InnovationMichael G. K. Jones (Supervisor) - Murdoch University, Centre for Crop and Food InnovationDean Diepeveen (Supervisor) - Murdoch University, School of Information TechnologyDavid Murray (Supervisor) - Murdoch University, School of Information Technology
- Awarding Institution
- Murdoch University; Doctor of Philosophy (PhD)
- Identifiers
- 991005723866307891
- Murdoch Affiliation
- School of Information Technology
- Resource Type
- Doctoral Thesis
Metrics
207 File views/ downloads
193 Record Views