Range-based Relative Localization Approach with Inexact Initial Coordinates
Liyuan Xu, Jie He, Peng Wang, Qin Wang, University of Science and Technology Beijing, China
Non-linear minimum optimization methods have been introduced in range-based relative localization problems in plenty of research. Despite this, common optimization algorithms, like maximum likelihood estimation method, depend heavily on initial coordinates to find an optimal solution. In contrast to open areas that Global Navigation Satellite Systems help provide these initial values for location, in areas with poor satellite signal reception, no exact initial value is available. This has led to inapplicability of the common optimization methods to non-linear localization pattern in these areas. As the inexact initial coordinate means infeasible solution to the optimization problem, accurate localization with these values is possible if the algorithm can find an optimal solution from infeasible solutions, and penalize mechanism is such a technique. This article presents a localization approach that performs accurate location estimation with random initial coordinates. Meanwhile, we validate the feasibility and efficiency of this approach.