Monte Carlo localization algorithm for indoor positioning using Bluetooth low energy devices

Abstract

This paper presents a technique for indoor localization using the Monte Carlo localization (MCL) algorithm. The MCL was upgraded from Markov localization, with both belonging to the family of probabilistic approaches. Throughout the last decade, laser rangefinders and gyroscopes have been applied to MCL-based robotic localization systems with remarkable success. However, the utilization of MCL-based indoor localization for mobile devices, by applying Bluetooth low energy (BLE) sensors, is still being researched. In this paper, we present a technique that utilizes MCL that exploits two sensors, namely, the accelerometer and compass, with commonly deployed BLE beacons to localize people with mobile devices indoors. Experimental results illustrate that, by applying MCL with BLE beacons using an accelerometer and compass, the error of the calculated coordinates for the user position is less than 1 m in line of-sight (LOS) environments, while in a complex non-LOS environment, the average error is 3 m. Meanwhile, the proposed MCL system does not demand a high deployment density of BLE beacons compared with triangulation and trilateration-based indoor positioning algorithms.