El sistema se prob\u00f3 en el laboratorio de GMV en el Reino Unido contra conjuntos de datos simulados y TEXBAT (Texas Spoofing Test Battery). El conjunto de datos TEXBAT es una colecci\u00f3n disponible p\u00fablicamente de escenarios de suplantaci\u00f3n de identidad grabados para evaluar las t\u00e9cnicas de autenticaci\u00f3n GNSS. Se llev\u00f3 a cabo una campa\u00f1a de prueba y validaci\u00f3n separada en las instalaciones ESTEC de la ESA en los Pa\u00edses Bajos, utilizando los datos recopilados en el evento Jammertest 2024 en Noruega, que implic\u00f3 una serie de ataques de suplantaci\u00f3n de identidad e interferencia por etapas.<\/p>\n
The European Space Agency (ESA)-funded BREGO project (\u2018Block-box for an optimized GNSS spectrum monitoring network using AI\u2019), carried out by Spanish technology company GMV, has developed a new system for real-time jamming and spoofing detection, classification and mitigation.<\/p>\n
The system employs optimized signal processing techniques driven by artificial intelligence (AI) and machine learning (ML) algorithms.<\/p>\n
The vulnerability of GNSS technologies is well known. A relatively low-power signal means GNSS signals can be nullified by jamming, while the openness of the signal structures, specifically GPS L1 and E1, make the technology vulnerable to spoofing, wherein false signals are substituted for genuine ones.<\/p>\n
Speaking at a recent presentation hosted by ESA, GMV GNSS Engineer Wahyudin Syam explained the rationale behind the BREGO project: \u201cGNSS receivers are deployed around the world and in many domains, providing position, velocity and time information for safety-critical, liability-critical, commercial and other applications. Our aim is to provide resilient navigation in environments dominated by GNSS threats.<\/p>\n
\u201cWe started with a review of the state of the art for GNSS threats, detection and mitigation techniques, and we considered technical specifications and the various trade-offs.\u201d A set of AI, ML, digital signal processing-based algorithms were modeled and tested, after an in-house dataset was used to train the algorithms for jamming and spoofing detection and classification.<\/p>\n
Solid work<\/strong><\/p>\n BREGO created a \u2018block-box\u2019 hardware setup, comprising an external radio frequency-to-radio frequency device for enhancing any consumer-grade, off-the-shelf GNSS receiver. Key system elements include a high-performance PC with a USRP X410, multi-channel, software-defined radio front end, and a Septentrio GPS L1 and Galileo E1 receiver.<\/p>\n The system was tested at GMV\u2019s UK laboratory against simulated and TEXBAT (Texas Spoofing Test Battery) datasets. The TEXBAT dataset is a publicly available collection of recorded spoofing scenarios for evaluating GNSS authentication techniques. A separate testing and validation campaign was carried out at ESA\u2019s ESTEC facility in the Netherlands, using data gathered at the 2024 Jammertest event in Norway, which involved a series of staged spoofing and jamming attacks.<\/p>\n \u201cThe block-box system is receiver agnostic and its effectiveness has been validated,\u201d said Syam. \u201cIt is flexible and configurable and provides effective interference mitigation for both chirp and CDMA-based interference attacks in real-time, laying the groundwork for broader GNSS applications.\u201d The project also developed a variant of an adaptive notch filter, mitigating rapidly changing interference without causing nonlinear phase distortions.<\/p>\n The BREGO project was funded under ESA\u2019s NAVISP program, supporting technology innovation in the European PNT sector.<\/p>\n