The vulnerability of the Global Navigation Satellite System (GNSS) open service signals to spoofing and meaconing poses a risk to the users of safety-of-life applications. This risk consists of using manipulated GNSS data for generating a position-velocity-timing solution without the user’s system being aware, resulting in presented hazardous misleading information and signal integrity deterioration without an alarm being triggered. Among the number of proposed spoofing detection and mitigation techniques applied at different stages of the signal processing, we present a method for the cross-correlation monitoring of multiple and statistically significant GNSS observables and measurements that serve as an input for the supervised machine learning detection of potentially spoofed or meaconed GNSS signals. The results of two experiments are presented, in which laboratory-generated spoofing signals are used for training and verification within itself, while two different real-world spoofing and meaconing datasets were used for the validation of the supervised machine learning algorithms for the detection of the GNSS spoofing and meaconing.

Global Navigation Satellite System (GNSS) meaconing and spoofing are being considered as the key threats to the Safety-of-Life (SoL) applications that mostly rely upon the use of open service (OS) signals without signal or data-level protection. While a number of pre and post correlation techniques have been proposed so far, possible utilization of the supervised machine learning algorithms to detect GNSS meaconing and spoofing is currently being examined. One of the supervised machine learning algorithms, the Support Vector Machine classification (C-SVM), is proposed for utilization at the GNSS receiver level due to fact that at that stage of signal processing, a number of measurements and observables exists. It is possible to establish the correlation pattern among those GNSS measurements and observables and monitor it with use of the C-SVM classification, the results of which we present in this paper. By adding the real-world spoofing and meaconing datasets to the laboratory-generated spoofing datasets at the training stage of the C-SVM, we complement the experiments and results obtained in Part I of this paper, where the training was conducted solely with the use of laboratory-generated spoofing datasets. In two experiments presented in this paper, the C-SVM algorithm was cross-fed with the real-world meaconing and spoofing datasets, such that the meaconing addition to the training was validated by the spoofing dataset, and vice versa. The comparative analysis of all four experiments presented in this paper shows promising results in two aspects: (i) the added value of the training dataset enrichment seems to be relevant for real-world GNSS signal manipulation attempt detection and (ii) the C-SVM-based approach seems to be promising for GNSS signal manipulation attempt detection, as well as in the context of potential federated learning applications.