A reliable range-free indoor localization method for mobile robots

Abstract

This paper presents an indoor localization method for the mobile robots using the radio signal values and the geometry path on the map. It is divided into three parts, the building the maps, the estimating the position of the robots, and the experiment. First, as the map generation, we build two kinds of the maps which are a radio-based map and a topological map. The radio-based map consists of the radio signals of the Wi-Fi access points at the certain locations. The geometry path map is composed of the topological data that the mobile robots can make as the motion trajectories. Second, as the localization process, we use the fingerprint estimation and the multi-hypothesis tracking (MHT) method. The fingerprint estimation can find the absolute area where the robot is placed by matching the radio-based map and the online observations of the Wi-Fi signals. The MHT method can obtain the optimal position in the topological map by considering the dead-reckoning motion. Finally, when it comes to the experiment, we performed the proposed localization method with a real mobile robot in an office environment to verify its effectiveness and performance. The experimental results show that the proposed method can estimate the accurate position of the robots even though they do not use the expensive range sensors such as the laser range finder.