Inaudible echo-based indoor position recognition (part 2)

Abstract

Indoor localization using mobile application has seen raised in demand in the market. Room Recognition using inaudible echo is one of the techniques for indoor localization. In addition, it is susceptible to phone orientation which had detrimental effect on room recognition accuracy. This project studies various methods to sense phone orientation in real-time to mitigate the impact of phone orientation on indoor position recognition. In this paper, we explore the characteristics of 3 orientation sensing sensors: accelerometer, gyroscope, and magnetometer. Each of the sensors is not perfect in determining the phone orientation but have certain benefits to compensate one another. The benefits of the 3 sensors will be combined using a sensor fusion algorithm to produce the best possible orientation estimation. While there are several ways to conduct sensor fusion, we will concentrate on using complementary filters in this project. Keywords: Indoor localization, Room recognition, Inaudible echo, Smartphone, Sensors, Phone orientation, Accelerometer, Gyroscope, Magnetometer, Sensor fusion, Complementary filter