GNSS Signal Correlation and Measurement Datasets for Interference Classification

GNSS datasets are of remarkable importance for the GNSS scientific community to improve the performance of receivers. The aim of this paper is twofold. Firstly, we describe a unique dataset, which contains a large variety of interferences and distortions such as jamming, spoofing, multipath, and Evil WaveForm (EWF). They have been collected from tests with live GNSS signals, except for the EWF signals that have been generated with a GNSS RF simulator. The data is available in different formats such as IQ samples, receiver output files, and correlator outputs. Secondly, we present an example of the application of the dataset based on machine learning classification. The goal is to assess the feasibility of classifying the received signals and the environment of the GNSS receiver by analyzing GNSS observables and correlator outputs from two satellites. The results reveal that machine learning algorithms perform promisingly as they can accurately identify GNSS signals.