Crowd Sensing for Public Transport Terminals Using Bluetooth Low Energy (BLE) Beacons

Current crowdsensing techniques are limited to two-dimensional positioning, which restricts their accuracy and applicability in complex environments. This paper introduces a novel three-dimensional (3D) system for object positioning, utilizing spline curve fitting to predict and validate position calculations. By leveraging a sufficient number of data points, spline curve fitting creates a smooth curve that accurately represents the trajectory of a crowd of objects in a generalized state. The methodology utilizes a Kalman filter to mitigate the high fluctuation of RSSI signals, enhancing the prediction of the locations of BLE devices. This approach also accounts for the number of devices in a 3D area, incorporating height and the distance between devices into the calculations. Based on the RSSI signals, the system estimates distances more precisely. The results demonstrate a significant improvement, with an average accuracy increase of 17% in the final position predictions. This advancement in 3D crowdsensing provides a more accurate and reliable framework for tracking and positioning objects in varied environments.