A map-growing localization algorithm for ad-hoc sensor networks.

摘要

The goal of this thesis research is to create a new algorithm for sensor network localization. Localization is an important process in deploying ad-hoc wireless sensor networks. Several localization algorithms have been developed. However, they do not achieve satisfactory performance on irregularly-shaped networks. In this thesis, I present a localization algorithm based on an idea of growing local maps. The simulation results show that this algorithm achieves better performance for well-connected but irregularly-shaped networks or networks with obstacles blocking connections between a node and some other nodes within its radio range. Compared with the popular localization algorithm APS [1], the map-growing algorithm is more accurate for C-shaped grid or hexagon networks with acceptable coverage (>%75) when the range error is less than or equal to 5%R, where R is the radio range.