Autonomous positioning techniques based on Cramér-Rao lower bound analysis

Abstract

We consider the problem of autonomously locating a number ofasynchronous sensor nodes in a wireless network. A strong focuslies on reducing the processing resources needed to solve therelative positioning problem, an issue of great interest inresource-constrained wireless sensor networks. In the first partof the paper, based on a well-known derivation of the Cramér-Raolower bound for the asynchronous sensor positioning problem, weare able to construct optimal preprocessing methods for sensorclock-offset cancellation. A cancellation of unknown clock-offsetsfrom the asynchronous positioning problem reduces processingrequirements, and, under certain reasonable assumptions, allowsfor statistically efficient distributed positioning algorithms. Cramér-Rao lower bound theory may also be used for estimatingthe performance of a positioning algorithm. In the second part ofthis paper, we exploit this property in developing a distributedalgorithm, where the global positioning problem is solvedsuboptimally, using a divide-and-conquer approach of lowcomplexity. The performance of this suboptimal algorithm isevaluated through computer simulation, and compared to previouslypublished algorithms. Copyright © 2006 Hindawai Publication Corporation. All rights reserved.