De-Biao Lu

The escalating prevalence of spoofing attacks with active manipulation intentions poses severe challenges in ensuring the security and trustworthiness of localization using Global Navigation Satellite System (GNSS). Many non-GNSS sensors that are immune to GNSS spoofing have been considered in multi-sensor structures to compensate the unavailability of the GNSS measurements. However, current sensor fusion solutions are not specifically designed to ensure the performance of localization without sacrificing security. In this paper, a Spoofing-resilient Factor Graph Localization (SRFGL) framework is designed to integrate multi-sensor measurements to defend against GNSS spoofing attacks while simultaneously recovering unbiased measurements and trustworthy positioning solutions. Specific countermeasures, including the adaptive spoofing detection and identification, spoofing mitigation based on spoofing profile estimation, and adaptive error modeling, are embedded within a whole factor graph structure to enhance the capability of state recovery against the intrusion by spoofing interference. Results using practically collected spoofing-scenario datasets demonstrate that the proposed solution can achieve spoofing-resilient localization under the multi-sensor integration architecture. Moreover, the flexibility of the proposed structure makes it easy to be seamlessly integrated into different GNSS-based localization instruments and equipment, contributing to proactive protection against the risk raised by GNSS spoofing.

The spoofing attack brings more serious threats and challenges to the Global Navigation Satellite System (GNSS) receiver. The rapid and accurate spoofing detection mechanism is of great significance to the credibility and security of GNSS-enabled transport applications. In this paper, an unsupervised classification solution is proposed to detect GNSS spoofing by analyzing the features of Coarse Acquisition (C/A) code Autocorrelation Function (ACF) using a Hybrid Convolutional Autoencoder (HCAE) method integrated with an attention-driven memory network. A dynamic threshold-based protection mechanism is introduced to reduce the system’s sensitivity to unexpected anomalies, thereby enhancing detection accuracy. The effectiveness of the proposed solution is verified by comparison with referencing detection methods using the Texas Spoofing Test Battery (TEXBAT) and spoofing injection test datasets. Specifically, the performance indices of the proposed method are improved over the involved referencing methods, which demonstrate that this solution can realize accurate and efficient detection of GNSS spoofing under the data-driven scheme.

The vulnerability of Global Navigation Satellite System (GNSS) may result in great risks and insufficiencies under the threaten of a GNSS spoofing attack, which constrains its utilization in safety-related applications, such as rail signaling and collision avoidance. An important strategy for safely utilizing GNSS in the rail vehicle system is the detection and identification of the spoofing attack status. In this work, we present an advanced spoofing detection solution for GNSS-based rail vehicle positioning, which consists of the signal parameter estimation phase and the spoofing detection phase. To determine the composition of the satellite signal, an estimation method is proposed to decompose the correlation function of the signal by combining the least absolute shrinkage and selection operator (LASSO) and least squares (LS). The optimal signal parameters of the two subsignals are determined, including signal amplitude and code delay. Three spoofing detection indicators are designed according to the relative change relationship of estimated parameters. For a difficult-to-detect synchronous-induced spoofing attack, a model-based unsupervised GNSS spoofing detection solution is proposed using a novel Bayesian optimized support vector data description (BOSVDD) method. The model integrates the characteristics of the three detection indicators to determine whether the system is attacked by spoofing. We apply our proposed solution under specific GNSS spoofing scenarios using the experimental results from a GNSS spoofing-injection test platform. Furthermore, we show the performance benefit of our solution compared to state-of-the-art parameter estimation strategies and spoofing detection methods.

As the positioning technology advances, the Global Navigation Satellite System (GNSS) has been highly concerned in railway systems. Meanwhile, the attack surface to GNSS-enabled train positioning has expanded, and thus the concerns regarding the positioning security have increased. To deal with potential threats from a GNSS spoofing attack, research on active detection is conducted. A framework of multi-domain feature-based spoofing detection is established, under which the semi-supervised GANomaly network is designed to generate the data-driven model. Different blocks, including the Multi-scale Group Dilated Convolution (MGDC), the Attention Gate (AG) and Double-dilated TCN Residual Block (DTCN ResBlock), are involved to improve the reconstruction ability of the vanilla GANomaly network. A dynamic threshold mechanism is adopted in identifying the existence of the spoofing attack. Experimental results with the open-source TEXBAT dataset under the ds7 static scenario and a railway train dataset from the spoofing injection test demonstrate the certain advantages of the proposed solution over the reference single-feature-based methods and existing networks, which highlight the potential for achieving the resilient train positioning under the observing conditions with respect to the adopted datasets.

Jamming to satellite navigation signals has become a major threat to the safety critical train control systems using Global Navigation Satellite System (GNSS), where the Virtual Balise (VB) concept enables a specification-compatible solution to utilize GNSS and reduce the physical Balises. This paper presents a novel GNSS jamming detection and exclusion solution to achieve trustworthy capture of VBs. An advanced architecture of Virtual Balise Reader (VBR) is proposed by integrating an Interference Qualifier Module (IQM) into the conventional GNSS-enabled VB framework. To enhance the trustworthiness of the VB capture, the IQM detects and identifies the existence of interference by examining the residuals between the real and predicted GNSS pseudo-ranges, which are generated by the odometry data and the trackmap. A discrimination test approach is utilized to evaluate the availability of raw satellite measurements according to the statistical analysis. We embed the performance indicator for each pseudo-range into the VB capture logic to isolate the degraded measurements and, meantime, adopt an adaptive Capture Interval Limit (CIL) to avoid the unexpected missed capture. Data sets from a GNSS jamming injection-based test platform are used for comparative studies, demonstrating the necessity and superiority of the proposed solution in improving the interference protection capability and trustworthiness of VB capture under the GNSS jamming environment.

In the frame of the development of artificial intelligent (AI) based validation tool for Global Navigation Satellite Systems (GNSS) localisation systems MATLAB Simulink models for all the developed AI-based methodologies have been created. The combination of Artificial Neural Network (ANN) based validation tools with the Mahalanobis Ellipses Filter (MEF) methodology for accuracy-based data evaluation for quantitative and qualitative analysis of trueness and precision, as well as Particle Filter (PF) techniques for position estimation to aid the reference system result in significant improvements for GNSS-aided localisation system, allowing the construction of a GNSS-dependent reference system for when no independent reference is available. These additional elements are the bases for future intelligent GNSS-based localisation systems with both quantitative and qualitative validation tools. The proposed prototype constructed within System functions (S-Functions) for extension of MATLAB Simulink capabilities presents total interconnectivity in real-time, implemented using commodity microcomputers and providing an interface between the MATLAB Simulink implemented algorithms and the real world. In the presented prototype a Raspberry Pi Model B microcomputer running a modified Debian Linux distribution provides General Purpose and UART ports that enable the implementation of the low-level communication driver for the GNSS receiver. The prototype enables real-time application of the algorithms. The prototype includes a GNSS receiver, with a sampling position rate and constellation data at 1Hz. NMEA format provides the necessary data to be processed with the MATLAB Simulink model of the ANN-based validation tools and the PF-based estimator. All algorithms are executed in a Windows PC and the communication between the microcomputer and the MATLAB Simulink model is achieved via TCP Sockets, with interface S-Functions to handle the communication. These methodologies will enable future safety-relevant applications, such as on-board uncertainty evaluation; advanced driver assistance systems; and GNSS-based vehicle localisation with intelligent maps for track selective.