A Method of Detections’ Fusion for GNSS Anti-Spoofing

Abstract

The spoofing attack is one of the security threats of systems depending on the Global Navigation Satellite System (GNSS). There have been many GNSS spoofing detection methods, and each of them focuses on a characteristic of the GNSS signal or a measurement that the receiver has obtained. The method based on a single detector is insufficient against spoofing attacks in some scenarios. How to fuse multiple detections together is a problem that concerns the performance of GNSS anti-spoofing. Scholars have put forward a model to fuse different detection results based on the Dempster-Shafer theory (DST) of evidence combination. However, there are some problems in the application. The main challenge is the valuation of the belief function, which is a key issue in DST. This paper proposes a practical method of detections’ fusion based on an approach to assign the belief function for spoofing detections. The frame of discernment is simplified, and the hard decision of hypothesis testing is replaced by the soft decision; then, the belief functions for some detections can be evaluated. The method is discussed in detail, and a performance evaluation is provided, as well. Detections’ fusion reduces false alarms of detection and makes the result more reliable. Experimental results based on public test datasets demonstrate the performance of the proposed method.

Detection of GNSS Spoofing by a Receiver in Space via Fusion of Consistency Metrics

“…Each consistency check provides its own soft output that can then be combined to return a unique hard output as a flag stating that a potential threat is present. The idea is to use a method based on Dempster-Shafer theory (DST) to fuse multiple detectors as done in [15].…”

Situational Awareness: Mapping Interference Sources in Real-Time Using a Smartphone App

“…Pre–correlation time-frequency (TF) domain analysis, mainly focusing on jamming signals, is adopted in [37,38,39]. Targeting specifically spoofing detection, works in [40,41] propose post–correlation approaches based respectively on a new particle filter for the positioning computation and detections’ fusion based on correlators output monitoring and Doppler consistency check. A subspace projection-based spoofing mitigation algorithm, relying on code delay and Doppler information, is detailed in [42].…”

A GPS Spoofing Generator Using an Open Sourced Vector Tracking-Based Receiver

“…In the signal features method, the features of the spoofing signal are quite similar to those of the actual signal and there is no sudden change in the transition process; but still, the signal feature method has not proven to work well. In addition, some crossing methods were proposed to detect spoofing, for instance machine learning [32], maximum likelihood estimation [33], and cooperation of multiple detections [34]. However, these methods are still dependent on prior information or actual signal features [35].…”