WSN Localization Using RSS in Three-Dimensional Space—A Geometric Method With Closed-Form Solution

Abstract

Proposed in this paper is a wireless sensor network source localization algorithm in the three-dimensional space. By using the measurements of the received signal strength (RSS), two closed-form none-ambiguous estimators are proposed to geometrically locate the source position under the shadowing noise in the wireless channel. The dynamic adjustment scheme has been applied in the localization algorithm to combat the measurements error in the ranging. With multiple measurements of the RSS, a lower bound of the source localization error is also derived. The numerical simulations show that the proposed estimators outperform the conventional least squares estimator, especially under the adverse scenario when the reference sensor nodes are close to be coplanar. Comparing with the maximum likelihood estimator that approaches the localization lower bound, the proposed estimators result in significantly lower computational burden, but without showing much degradation of localization accuracy.