Indoor Tracking System using Bluetooth Low Energy

Tan, Khai Shean (2021) Indoor Tracking System using Bluetooth Low Energy. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Tan Khai Shean.pdf Restricted to Registered users only Download (3MB)

Abstract

Location tracking technologies has experienced a rapid development in many sectors and has revolutionized human daily life in recent years such as the Global Positioning System (GPS). Many GPS based apps has been introduced to the public such as Google maps, Grab, Waze and other app which widely used in our daily life. However current location tracking technologies are just only applicable for outdoor environment. Therefore, indoor positioning system (IPS) has been an active research fields by the researcher to overcome the weakness of GPS. In this research is to develop an indoor tracking system which applied with Bluetooth Low Energy (BLE) devices, Received Signal Strength Indicator (RSSI) Fingerprinting, and Convolutional Neural Network (CNN) to perform location tracking. The objective for this research is to evaluate the tracking accuracy of the CNN model when applied with different RSSI fingerprint that obtained though BLE beacon. In a nutshell, the accuracy obtained by the CNN model for location tracking is 91.18% and the acceptability for the wrong result has also achieved for 66.67%. However, the indoor tracking system developed still can be improved by reducing the fluctuation of RSSI during fingerprinting which will also reduce the latency at the same time.