In the beginning of the 21th century, the goal of telecommunications was to connect everyone anytime anywhere in the world. Now, with the tremendous developments in electronics and communications, we are thinking to go beyond our prior goal and connect anything anytime anywhere (Internet of things). This dream would not have been shaped without the recent advances in wireless sensor networks, and the exploding use of this technology in our everyday life. Still to fulfill such a purpose, we have to face physical and technological challenges especially when it comes to outer space or underwater. In this thesis, we study the fundamental problem of underwater sensor node localization as an indispensable task for any network.
In the first part of this thesis, the main challenges of underwater acoustic communications are reviewed, and their effects on underwater localization algorithms are discussed. It is shown how these algorithms can be categorized into different groups and how they can be evaluated through several metrics.
The second part of the thesis focuses on the development of accurate localization algorithms in an underwater medium with a variable sound speed profile (SSP). The SSP is modeled by linear and piecewise linear functions of the water depth, and optimal localization algorithms are proposed. It is shown that for large network sizes, the performance of the proposed algorithms is much better than traditional approaches.
The large propagation delay of acoustic waves plays an important role in the required time to perform node localization. In order to minimize the localization time, two packet scheduling schemes are proposed in the third part of this thesis, collision-free and collision-tolerant schemes. In this part, through a detailed analysis and numerical results, we show how a localization algorithm can benefit from optimal medium access control design.