Node Self-Localization Algorithms For 3D Wireless Sensor Networks

As a new method to collect and process information, wireless sensor networks (WSNs) will have a wide range of applications. Gathered data are meaningless if they are not combined with the positions of sensors, so node self-localization becomes one of the supporting techniques for WSNs.There are three fundamental node self-localization algorithms for WSNs, i.e., Centroid, DV-Hop, and DV-Distance, which are the bases of many improved algorithms. However, these proposed algorithms only focus on node self-localization for 2D networks. Practical nodes of WSNs are generally arranged in 3D scenarios such as in space, under water, or on hilly terrains, rather than on pure 2D fiats, thus 3D position information of nodes is commonly required. Node self-localization algorithms for 3D WSNs are mainly researched in this thesis.Based on 2D localization algorithms, i.e., Centroid, DV-Hop, and DV-Distance, the corresponding 3D localization algorithms, i.e., 3D-Centroid, 3D-DV-Hop, and 3D-DV-Distance are proposed in this thesis. Average localization error, coverage, and the number of messages exchanged are evaluated as the performance critera for localization algorithms with a C++ simulation platform. For each pair of localization algorithms, firstly the performance of the 2D algorithm on 2D plats is presented, and then the performance of the 2D algorithm and the 3D one in 3D spaces is compared on comparable conditions.The work of this thesis indicates that, in 3D scenarios, the 2D localization error, the coverage and the number of exchanged messages of 3D-Centroid are the same as those of Centroid. However, compared with corresponding 2D localization algorithms, 3D-DV-Hop and 3D-DV-Distance gain much less 2D localization error, slightly less coverage, and the same number of messages exchanged. The above fact implicates that even if we do not need to know the height information of nodes, it is still necessary to employ 3D localization algorithms in 3D scenarios. Apparently, if we need the third dimensional location information in addition to 2D positons, then 3D localization algorithms are indispensable. Meanwhile, the implementations of 3D-Centroid, 3D-DV-Hop, and 3D-DV-Distance are similar to those with their corresponding 2D localization algorithms, i.e., Centroid, DV-Hop, and DV-Distance.