Classification of GNSS Jammers using Machine Learning : Multivariate Time Series and Image Classification Based Approaches
MetadataShow full item record
Original versionVoigt, J.M. (2021) Classification of GNSS Jammers using Machine Learning : Multivariate Time Series and Image Classification Based Approaches (Master's thesis). University of Agder, Grimstad.
GNSS is one of the most widely used positioning techniques in modern technology. However, the signals from the GNSS satellites are weak on earth due to the propagation attenuation, making GNSS-based systems vulnerable to interference or jamming of signals. This the-sis proposes a machine learning approach to detect the presence of jammers in the GNSS spectrum bands in recorded data. We have employed the state-of-the-art and baseline machine learning techniques for Image- and multivariate time series classification and evaluated their ability to classify the presence of illegal jammer activity. In addition, we propose a novel complexity reduced version of a recently proposed multivariate time series transformer model. Experiment results show that the tested machine learning techniques, after proper configurations, achieve a classification accuracy of up to 99.5%. Moreover, the simplified transformer-based approach achieves the same level of performance while reducing the number of parameters by nearly half compared to comparable artificial neural network models. The high accuracy confirms the applicability of the machine learning approach in jammer classification.
Master's thesis in Information- and communication technology (IKT590)