Improving Performance of Bluetooth Low Energy-Based Localization System Using Proximity Sensors and Far-Infrared Thermal Sensor Arrays

This paper presents the concept of a hybrid positioning scheme using results from a Bluetooth Low Energy (BLE)-based system and additional infrared (IR) devices: proximity sensors and far-infrared thermal sensor arrays. In the proposed solution, the IR sensors operate independently from the BLE subsystem. Their output (the distance to the localized person and the angle between the sensor axis and the person's location) is periodically used to improve the positioning accuracy. The results from both parts of the system are fused using a particle-filter-based algorithm. The proposed concept was tested experimentally. The initial tests established that both the proximity (VL53L5CX) and array (MLX90640) sensors allowed for angle estimations with a mean accuracy of about a few degrees. Using them in the proposed hybrid localization scheme resulted in a mean positioning error decrease of several centimeters.