A Novel Range-Selective Generalized Weighted Centroid Method for Source Localization in Bearing-Only Sensor Networks

In this study, the source localization problem is investigated using a group of networked sensors if only bearing measurements are available. The existing bearing-only source localization (BOSL) methods are subjected to a critical problem, i.e., how to achieve a fine balance between the computational complexity and the positioning accuracy when the threshold effect is considered. A novel range-selective generalized weighted centroid localization (RSGWCL) estimator is proposed to address the above-mentioned problem. Without the threshold effect, the proposed RSGWCL method is capable of achieving comparable positioning accuracy to the maximum likelihood estimator (MLE). If the threshold effect is identified, the RSGWCL estimator still has the capability of providing reliable estimates and outperforms the MLE. The result of the simulation is analyzed, such that the performance of the proposed method can be verified.