A Device-free Indoor Localization System Based on Supervised Learning and Bluetooth Low Energy

Abstract

Passive, device-free indoor localization is a research field that has drawn the attention of many researchers in the last years. A reliable, low-cost and accurate solution that does not require an active device at the target has not been found yet. In this work, we tackle this challenge by proposing a pattern recognition method based on supervised learning models. This method classifies passive fingerprints that represent the influence of a target’s position on the field strength distribution into possible locations within the test area. The method leverages the ability of an optimized custom BLE sensor platform for IoT applications. A smart feature extraction is then proposed and the data is fed to supervised machine learning models such as KNN and SVM. It will be shown based on experimental results and descriptive statistics that passive localization of a target can be carried out robustly and the median error of the proposed method is 0.67 m.