Robust Multidimensional Scaling for Cognitive Radio Network Localization

Abstract

Localization of primary users (PUs) and secondary users (SUs) is one of the essential features of cognitive radio networks (CRNs). Given that there is no communication between PUs and SUs, localization of the whole network is a challenging task. In this paper, we propose a two-stage localization algorithm that combines multidimensional scaling (MDS) and Procrustes analysis for a CRN with proximity information. In the proposed algorithm, a hybrid-connectivity-and-estimated-distance-based strategy is introduced to get maximum benefit from the information available in the network. Simulations are included to compare the proposed algorithm with weighted centroid localization (WCL) in terms of the root-mean-square-error (RMSE) performance, as well as the Cramer-Rao lower bound (CRLB) for CRN localization. It is proved that the proposed algorithm outperforms the WCL solutions for the CRN localization problem.