Mobile node-aided localization and tracking in terrestrial and underwater networks

Gong, Zijun (2021) Mobile node-aided localization and tracking in terrestrial and underwater networks. Doctoral (PhD) thesis, Memorial University of Newfoundland.

[English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (4MB)

Abstract

In large-scale wireless sensor networks (WSNs), the position information of individual sensors is very important for many applications. Generally, there are a small number of position-aware nodes, referred to as the anchors. Every other node can estimate its distances to the surrounding anchors, and then employ trilateration or triangulation for self-localization. Such a system is easy to implement, and thus popular for both terrestrial and underwater applications, but it suffers from some major drawbacks. First, the density of the anchors is generally very low due to economical considerations, leading to poor localization accuracy. Secondly, the energy and bandwidth consumptions of such systems are quite significant. Last but not the least, the scalability of a network based on fixed anchors is not good. Therefore, whenever the network expands, more anchors should be deployed to guarantee the required performance. Apart from these general challenges, both terrestrial and underwater networks have their own specific ones. For example, realtime channel parameters are generally required for localization in terrestrial WSNs. For underwater networks, the clock skew between the target sensor and the anchors must be considered. That is to say, time synchronization should be performed together with localization, which makes the problem complicated. An alternative approach is to employ mobile anchors to replace the fixed ones. For terrestrial networks, commercial drones and unmanned aerial vehicles (UAVs) are very good choices, while autonomous underwater vehicles (AUVs) can be used for underwater applications. Mobile anchors can move along a predefined trajectory and broadcast beacon signals. By listening to the messages, the other nodes in the network can localize themselves passively. This architecture has three major advantages: first, energy and bandwidth consumptions can be significantly reduced; secondly, the localization accuracy can be much improved with the increased number of virtual anchors, which can be boosted at negligible cost; thirdly, the coverage can be easily extended, which makes the solution and the network highly scalable. Motivated by this idea, this thesis investigates the mobile node-aided localization and tracking in large-scale WSNs. For both terrestrial and underwater WSNs, the system design, modeling, and performance analyses will be presented for various applications, including: (1) the drone-assisted localization in terrestrial networks; (2) the ToA-based underwater localization and time synchronization; (3) the Doppler-based underwater localization; (4) the underwater target detection and tracking based on the convolutional neural network and the fractional Fourier transform. In these applications, different challenges will present, and we will see how these challenges can be addressed by replacing the fixed anchors with mobile ones. Detailed mathematical models will be presented, and extensive simulation and experimental results will be provided to verify the theoretical results. Also, we will investigate the channel estimation for the fifth generation (5G) wireless communications. A pilot decontamination method will be presented for the massive multiple-input-multiple-output communications, and the data-aided channel tracking will be discussed for millimeter wave communications. We will see that the localization problem is highly coupled with the channel estimation in wireless communications.

Actions (login required)

View Item