스마트폰을 이용한 실시간 실내 위치 측위 시스템 구현 및 성능 분석

Abstract

The need of indoor positioning has been rapidly increasing with the advent of different location-based services. Considerable efforts have been made to find the user’s position by using wireless networks because of the low cost and high availability of wireless access. The fingerprinting positioning method based on received signal strength (RSS) pattern matching has gained popularity in indoor environment. Unfortunately, due to environmental factors including noise, co-channel interference, multipath, shadowing, difference of device type and others, RSS measurements remarkably vary although the user’s position is fixed. To mitigate the problems of RSS variation, this thesis proposes a fused inertial measurement unit (IMU)-assisted RSS correction and nearest neighbor selection (NSS) algorithm. The proposed algorithm estimates a RSS conversion characteristic using a predicted position between the measured RSS and stored RSS in database, and processes an IMU-assisted NSS algorithm that filters out irrelevant reference points based on the position prediction with IMU measurements. The proposed algorithm is implemented into an Android-based smartphones in a real environment. In this thesis, a range-free localization system using commercial smartphones is also presented with Bluetooth capabilities to investigate the feasibility of range-free localization for indoor positioning. In the range-free localization system, each smartphone periodically scans nearby Bluetooth enabled devices and sends the results to the localization server. The localization server collects the scanning results from smartphones for a short period and finds their positions via range-free algorithms. In addition to the localization server, a monitoring unit has been developed to supervise the range-free localization system. The range-free localization system also was tested in a real environment.