One of the main challenges of Moving Target Indication (MTI) with spaceborne radars is distinguishing moving targets from strong ground clutter, primarily due to the high velocity of the radar platform. This issue is particularly pronounced in spaceborne systems, where satellites in Low Earth Orbit (LEO) travel at speeds of approximately 7,000 m/s. Such rapid motion causes a broadening of the clutter Doppler spectrum, making it especially difficult to detect slowly moving targets, as their signals are often masked by the clutter. To address this challenge, advanced MTI techniques are required, which rely on multiple receiving antennas arranged along the flight direction and separated by a defined baseline. These multiple antennas enable spatial sampling, allowing the system to resolve differences not only in Doppler frequency but also in angle, thereby enhancing the ability to separate moving targets from ground clutter. Among these techniques, Space-Time Adaptive Processing (STAP) has been the most extensively studied in the literature, as it offers full adaptivity in both space and time domains.

In this study, we implement STAP and evaluate its efficiency in a SwarmSAR system, which is a multistatic spaceborne radar configuration where each satellite carries a single antenna and serves as an individual element of the distributed array. To the best of our knowledge, STAP has not previously been applied in such a configuration. The results demonstrate that STAP can be effectively applied, with clutter suppression improving as the target moves further from the clutter subspace. Due to the significantly larger baselines in SwarmSAR systems compared to the wavelength, targets can be detected at relatively low velocities (below 1 m/s). However, these large baselines are also the primary limitation to STAP performance in a SwarmSAR topology. They give rise to grating lobes, which cause angular ambiguities that, due to the coupling between angle of arrival and target velocity introduced by platform motion, translate into velocity ambiguities. Consequently, multiple angle-velocity pairs fall within the clutter subspace, making it impossible to detect targets with those specific combinations.

Moreover, when the baselines are large enough that each satellite measures a different target velocity—differing by more than the velocity resolution—the efficiency of STAP decreases by a factor equal to the array gain, and each satellite produces a separate detection. To address this, we propose a strategy to achieve the same observation geometry across all satellites. Results from this approach demonstrate that it is possible to recover the array gain and achieve a single detection per target. Additionally, when the baselines are non-uniform, the ambiguities in angle of arrival and target velocity are also resolved. It is important to note that these results were obtained under several ideal assumptions, including perfect synchronization and phase stability between satellites, precise knowledge of satellite positions, and stationary clutter.