Hello
Matteo Palescandolo
Ph.D. Student in Industrial Engineering, University of Naples Federico II
About Me
As low Earth orbit (LEO) becomes increasingly congested, traditional Space Situational Awareness (SSA) methods—such as ground-based radar and optical telescopes—face significant coverage gaps and operational constraints. To address these limitations, my research investigates space-based passive Radio Frequency (RF) tracking architectures to provide continuous, weather-independent localization of cooperative space objects. The core objective is to develop GNSS-resilient tracking systems by utilizing Time Difference of Arrival (TDOA) and Frequency Difference of Arrival (FDOA) measurements collected by observing LEO satellite constellations.
A significant portion of my work involves designing high-fidelity simulation environments to validate these tracking architectures. This framework simulates the entire operational pipeline, encompassing orbital propagation, dynamic Signal-to-Noise Ratio (SNR) evaluation, Earth occultation filtering, and synthetic measurement generation. To accurately resolve the non-linear localization problem, the research introduces a robust three-stage estimation algorithm combining multi-start initialization, Trust-Region-Reflective optimization with analytical Jacobians, and strict physical validation bounds.
The primary focus of this research is quantifying the critical relationship between constellation geometry and tracking performance. By employing the simulation framework to analyze four-satellite cluster architectures, the research demonstrates that the Geometric Dilution of Precision (GDOP) is the dominant factor determining localization accuracy. For instance, a geometrically diverse configuration combining co-orbital and cross-track sensors yielded mean position and velocity accuracies of 16.93 m and 0.80 m/s when GDOP remained below 5. Conversely, the study proved that strict co-orbital layouts lead to massive error amplification (errors exceeding 16 km), while symmetric cross-track geometries induce bimodal ambiguity, causing the solver to map targets on the wrong side of the orbital plane.
Ultimately, this research establishes critical design guidelines for future Space Traffic Management architectures, proving that spatial diversity in satellite deployment is essential for accurate and unambiguous RF-based localization.
Building upon this static localization framework, my current research advances to dynamic tracking by implementing a customized Extended Kalman Filter (EKF). To address realistic operational scenarios, the filter is initialized directly from existing orbital catalogs (e.g., Two-Line Elements) utilizing structured prior uncertainties mapped within the Radial-Transverse-Normal (RTN) frame. Furthermore, the EKF rigorously integrates J2 gravitational perturbations and the gravity gradient tensor to accurately model orbital dynamics. This ensures highly precise dead reckoning and robust track-loss recovery during measurement gaps caused by Earth occultation or degraded signal-to-noise ratios.
Click here to view my poster.
A significant portion of my work involves designing high-fidelity simulation environments to validate these tracking architectures. This framework simulates the entire operational pipeline, encompassing orbital propagation, dynamic Signal-to-Noise Ratio (SNR) evaluation, Earth occultation filtering, and synthetic measurement generation. To accurately resolve the non-linear localization problem, the research introduces a robust three-stage estimation algorithm combining multi-start initialization, Trust-Region-Reflective optimization with analytical Jacobians, and strict physical validation bounds.
The primary focus of this research is quantifying the critical relationship between constellation geometry and tracking performance. By employing the simulation framework to analyze four-satellite cluster architectures, the research demonstrates that the Geometric Dilution of Precision (GDOP) is the dominant factor determining localization accuracy. For instance, a geometrically diverse configuration combining co-orbital and cross-track sensors yielded mean position and velocity accuracies of 16.93 m and 0.80 m/s when GDOP remained below 5. Conversely, the study proved that strict co-orbital layouts lead to massive error amplification (errors exceeding 16 km), while symmetric cross-track geometries induce bimodal ambiguity, causing the solver to map targets on the wrong side of the orbital plane.
Ultimately, this research establishes critical design guidelines for future Space Traffic Management architectures, proving that spatial diversity in satellite deployment is essential for accurate and unambiguous RF-based localization.
Building upon this static localization framework, my current research advances to dynamic tracking by implementing a customized Extended Kalman Filter (EKF). To address realistic operational scenarios, the filter is initialized directly from existing orbital catalogs (e.g., Two-Line Elements) utilizing structured prior uncertainties mapped within the Radial-Transverse-Normal (RTN) frame. Furthermore, the EKF rigorously integrates J2 gravitational perturbations and the gravity gradient tensor to accurately model orbital dynamics. This ensures highly precise dead reckoning and robust track-loss recovery during measurement gaps caused by Earth occultation or degraded signal-to-noise ratios.
Click here to view my poster.