Distributed mission architectures have gained significant attention in recent years as a promising alternative to traditional monolithic spacecraft designs. Multi-satellite systems, typically composed of smaller platforms, enable improved mission resilience, scalability, and performance through cooperative operations. Among the various applications, Earth Observation (EO) has emerged as a compelling domain. A key technology in EO missions – that can benefit from closer-range cluster of satellites - is passive radiometry, which involves detecting the energy emitted or reflected by the Earth’s surface across the electromagnetic spectrum. Traditionally, these missions have relied on a single monolithic satellite, with measurement accuracy closely linked to the size of the onboard instrument. This performance can be enhanced by employing a cluster of satellites in a tight and stable configuration, acting as a virtual larger instrument.
Several distributed mission concepts for passive radiometry have been investigated at Politecnico di Milano, including the TriHex mission (Martín-Neira, et al., 2023). This concept comprises three satellites flying in a fixed triangular formation, with relative distances of approximately nine metres. No mission to date has achieved such a challenging architecture with more than two satellites, particularly when relying solely on GNSS measurements for relative navigation, as the TriHex mission is designed to do to minimise system complexity. Previous studies have demonstrated the feasibility of this mission through the successful development of guidance and control solutions (Scala, 2022). However, a comprehensive assessment of its feasibility from a relative navigation perspective remains lacking.
The short distances between satellites in formations such as TriHex impose stringent requirements on the navigation subsystem, which must be robust and autonomous to deliver accurate relative positioning under a variety of contingencies. Such autonomy can be achieved through an Adaptive Extended Kalman Filter (AEKF), in which the process and measurement covariance matrices are tuned in real time. This adaptive approach enhances filter robustness against modelling errors, measurement outages, and poor GNSS geometry, ensuring reliable relative navigation performance throughout the mission.
Previous studies have demonstrated that millimetre-level accuracy can be achieved in swarm missions using only measurements from multiple GNSS constellations processed through an Extended Kalman Filter (EKF), as shown by Giralo (2019). Separately, Fraser (2021) introduced adaptive EKF techniques for two-satellite formations, employing maximum likelihood estimation and fuzzy logic to improve robustness under uncertainties.
The objective of this work is to demonstrate the feasibility of GNSS-only navigation for close-proximity systems comprising more than two satellites, with relative distances on the order of metres. This is achieved by the development of an adaptive Kalman filter tailored for multi-satellite missions with at least three satellites. By doing so, this research fills a critical gap in the development of robust filtering strategies for GNSS-only relative navigation in formations with such tight baselines, bringing a step forward in the feasibility assessment of unprecedented distributed mission concepts for passive radiometry.
Moreover, the study investigates inter-satellite information exchange within TriHex, analysing what data should be shared, how it should be transmitted, and its impact on navigation accuracy and system robustness. These two aspects, adaptive filtering and communication strategy, are essential for developing a navigation system capable of guaranteeing feasibility and safety for satellite swarms.
The proposed algorithms are validated using high-fidelity GNSS measurement from GPS and Galileo, generated by GEMS, an in-house software simulator developed at Politecnico di Milano (Michelazzi, Gaias, & Colombo, 2025). GEMS acts as a digital twin of a hardware testbed, accurately reproducing signal errors and receiver noise, thereby providing a realistic environment for performance assessment.

References
Fraser, C. T., & Ulrich, S. (2021). Adaptive extended Kalman filtering strategies for spacecraft formation relative navigation. Acta Astronautica, 178, 700-721.
Giralo, V., & D'Amico, S. (2019). Distributed multi-GSS timing and localization for nanosatellites. Navigation, 729-746.
Martín-Neira, M., Scala, F., Zurita, A., Piera, M., Duesmann, B., Drusch, M., . . . Corbella, I. (2023). TriHex: Combining Formation Flying, General Circular Orbits, and Alias-Free Imaging, for High-Resolution L-Band Synthesis. IEEE Transactions on Geoscience and Remote Sensing, 61, 1-17.
?Michelazzi, A., Gaias, G., & Colombo, C. (2025). Modelling and Simulation of GNSS Observables for Spacecraft Navigation. 1st IFAC Workshop on Control Aspects of Multi-Satellite Systems. Würzburg, Germany.
Scala, F. (2022). Multi Satellites Formation Flying for Earth Observation Applications in Low Earth Orbits. PhD Thesis, Politecnico di Milano.