Cautious Greedy Strategy for Bearing-based Active Localization: Technical Report

Abstract

We study a novel sensing model which occurs in wildlife monitoring applications. The sensor can be approximated as a bearing-to-target sensor, but with several critical attributes that set it apart from those addressed in bearing-only tracking literature. We propose a measurement strategy that can guarantee reduction in uncertainty in competitive time, regardless of the problem instance. We provide theoretical analysis in the form of a competitive proof, numerical studies, and results from field trials which confirm the applicability of the algorithm.