A Filtering Method for Machine Learning Utilization of ADS-B Data

The development of low-cost aircraft surveillance systems based on Automatic Dependent Surveillance-Broadcast (ADS-B) technology has gained considerable interest, leading to many applications. Our interest is particularly in harnessing these data for developing flight prediction models and their applications considering the information in ADS-B signals. In this paper, we propose a filtering method and assess its effectiveness for processing ADS-B data to enhance the accuracy of predicted flight location coordinates. The evaluation results demonstrate that the proposed method can successfully separate single routes for Machine Learning purpose. Also, the predicted data are very close to learning data and the observed errors are quite small.