Sensor Network Localization in Constrained 3-D Spaces

Localization of sensor nodes is important for many sensor network applications such as distance-based routing and target tracking. This paper presents a localization approach for sensor networks in constrained 3-D space. By assuming that all the original beacon nodes are on the bottom plane, the localization procedure for the entire network is primarily from the bottom to the top and at the same time the localization process is carried out in all directions from the regions where the original beacon nodes are clustered. Simulated results showed that efficiency is improved for network structures whose profiles are cuboid or cone shaped, and whose node distributions are layered or random. The effects of the number of original beacon nodes as well as the node density on the localizing errors and/or the localizing successful rate are explored. Finally, a large scale sensor network is analyzed to verify the propagating trend of localization