Using Two Nodes and Neighbors Examining to Improve Positioning Capability in Sensor Networks

Traditional triangulation positioning requires that a node needs three position-known neighbors to be positioned. This requirement limits the positioning capability, or the number of nodes being positioned, particularly for sparse sensor networks. We propose a positioning approach, two nodes and neighbors examining, by which the node with less than three position-known neighbors may still be positioned as long as its surroundings provide enough useful information. Our simulation results show that the number of nodes being positioned may be dramatically improved. We also consider distance measurement errors which may cause triangulation fail. The two nodes and neighbors examining approach may be used to estimate positions. Besides, we discuss how to estimate positions for some special cases and refine the estimations, and then illustrate the simulation results