Intelligent fingerprint-assisted for Indoor Positioning System

In this study an Indoor Positioning System (IPS) based on intelligent fingerprint-assisted has been proposed in order to calculate the reliable reference point data and development a positioning system learning mechanism through wireless network connection. Due the change in external environment, the traditional statistical approach will require more than 150 times of learning for the reference points, while the method proposed in this study will only require 100 times of learning to achieve the learning objective. This indicates that the reference point learning has been accelerated, and the learning efficiency has been improved. By using three units of AP for positioning in the environment of this research, the average positioning error has been improved from 1.61m to 1.46m with the positioning accuracy enhancement of 9.3%, thus improving the user satisfaction and usage rate of positioning system.