New results on node localizable conditions for sensor networks

This paper studies the problem of localizable conditions for sensor nodes in a two-dimensional sensor network. A new group of conditions are proposed which judge each node's localizability through analyzing its connections with not only a set of already localized nodes but also its other neighbors. Our new conditions allow up to four nodes to be localized at a time. This algorithm offers a substantial improvement in localizability over the well-known trilateration method and convex hull method. We show that a newly developed WHEEL extension condition is a special case of our conditions. We also demonstrate that our result can be used as a guideline to modify topology of a network for better localizability