Bluetooth low energy (BLE) based asset tagging system

Abstract

Asset Tracking is a valuable technology that most businesses want to leverage on, especially the well developed GPS-based outdoor asset tracking system. However, indoor localization is still not well developed as GPS is not accurate in indoor environment. One good option is to utilize BLE technology for indoor localization. However, by using the RSSI value itself is not accurate. Therefore, this project will develop an asset tracking system to collect RSSI fingerprinting data and increase localization accuracy by using various machine learning algorithms. The experimental results show a significant improvement over a previous BLE indoor localization study, but there is still opportunity for improvement, such as adopting different machine learning techniques as a comparison.