One strategy for toughening receivers is direction-of-arrival sensing. The technique relies on a multi-element GNSS antenna or the equivalent. Such techniques are uniquely well suited to the detection and mitigation of jamming and spoofing attacks on land, air, and sea vehicles. We have examined and developed several multi-element technologies.

This project will follow two paths in parallel, both involving advanced Global Navigation Satellite System (GNSS) receiver signal processing. The first is focused on defending against spoofing, the second against jamming. 

While GNSS is the primary means to provide absolute position information in transportation systems, radio frequency methods to detect anomalous GNSS signals may not enable their exclusion in all events. The multiple sensors incorporated into advanced vehicles and transportation systems offer unique opportunities to combat nefarious activities such as GNSS jamming and spoofing to maintain PNT accuracy and integrity. This project will involve three research directions related to GNSS multi-sensor augmentation.  

GNSS is vulnerable to jamming because the power of GNSS signals received near the Earth’s surface is extremely weak, as low as a tenth of a millionth of a billionth of a Watt.  Higher power signals from Low-Earth Orbiting (LEO) satellites intended for communication can be used as an opportunistic means of navigation, but only with significant risks because the LEO service providers have no commitment to navigation users.  In contrast, recently-modernized and emerging dedicated LEO constellations can provide positioning navigation and timing (PNT) with quantifiable performance. In particular, CARNATIONS industrial partners Satelles, Inc. and Xona Space Systems, two PNT LEO satellite constellation operators, provide signals that are secure, powerful, reliable, and independent of GNSS.  This project aims at (1) designing LEO satellite-based resilient PNT (R-PNT) algorithms and (2) evaluating them for transportation applications. 

Enabling mission critical communication for vehicular networks can be achieved by exploiting 5G and 6G wireless systems. However, given that such systems are primarily designed with high rate services in mind (e.g., multimedia), ensuring continuous availability of the communication link for V2X communication is a major challenge. The goal of the first year of this project is to explore the use of resilience as a metric for guaranteeing the operation of V2X links under different dynamics of the environment. This will create a seed for the next year of this project that can exploit the developed fundamentals to investigate other avenues like integrated sensing and communications. 

This project will follow three paths in parallel, all focused on developing vehicle strategies that provide improved knowledge of and resilience to positioning uncertainty, in particular, of the potential risk of spoofing. The first path is focused on developing resilient connected and automated vehicle (CAV) applications given uncertain PNT services; the second is developing resilience techniques through a multi-agent community approach; and the third is to conduct research on collaborative radio-frequency interference (RFI) localization. 

This project will leverage Edge Artificial Intelligence (Edge AI) to enhance the resilience of Global Navigation Satellite Systems (GNSS) in challenging environments. By deploying AI algorithms and models on edge devices, the project aims to reduce reliance on cloud infrastructures, particularly in dense blockage scenarios where GPS signals are weak or disrupted. The goal is to explore how bringing intelligence to the edge node can improve GNSS performance and resiliency in these difficult conditions.

This project enhances jamming and spoofing detection capabilities in Connected and Autonomous Vehicle (CAV) applications by utilizing information broadcasted by surrounding vehicles and other signals of opportunity in various driving scenarios. The focus is on vehicle-to-vehicle maneuvers, where malicious actors attempt to disrupt cooperative CAV operations. The project involves regular meetings to exchange ideas, share insights, and review progress, with potential for co-authoring conference papers and journal publications. Additionally, ongoing collaboration with StarNav will provide real-world spoofing data to support the research.

This project builds end-to-end cooperative and remote driving pipelines using C-V2X radios and infrastructure connections to establish a foundation for resilience studies in autonomous driving. While autonomous driving has made significant progress in recent years, widespread adoption still faces challenges related to handling corner cases. These challenges arise from limited visibility for ego vehicles due to obstructions and conditions, despite a full suite of sensors, as well as complex or unforeseen scenarios that autonomous vehicles are not designed to handle. To address these issues, the project emphasizes the importance of cooperative driving, where connected vehicles share sensor data and information, and remote driving, where a remote human operator takes control of an autonomous vehicle when necessary. Both approaches require stable and secure vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) connections, along with stringent latency and high bandwidth, particularly for video and LiDAR data delivery and sharing.