Robust Rigid Body Localization Under Position Error

Abstract

Rigid body localization (RBL) technology, which can obtain both the position estimation and direction information of a target, has been incorporated into location based services (LBS). However, most existing works on the rigid body based LBS assume that there are no sensor position errors, which can lead to misleading results. This paper proposes a semi-definite relaxation method, in which both rigid body sensors and anchor nodes position errors are taken into account. Since the original problem is non-convex, we transform it into the auxiliary variable based semi-definite programming via squaring both sides of range measurement. Simulation results demonstrate that the proposed algorithm outperforms the existed works in terms of root mean square error (RMSE).