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

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 irregular networks. In this paper, we present a localization algorithm based on an idea of growing local maps. Simulation was performed on networks of different shapes with different connectivity and anchor. Compared with APS DV-distance (Niculescu and Nath, 2001), the map-growing algorithm is about two times more accurate for C-shaped grid or hexagon networks with the same coverage as of APS DV-distance, when the range error is less than or equal to 10%R, where R is the radio range. Compared with APS Euclidean (Niculescu and Nath, 2001), the coverage of map-growing is 100% with low GPS ratio while APS Euclidean requires a higher GPS ratio to get a better coverage.