The growth in the overpopulation of resident space objects calls for space surveillance initiatives. In particular, the threat posed by in-orbit collisions and fragmentations, as well as by satellites re-entry requires an efficient space objects cataloguing capability. Ground-based sensors are the main contributors to build up and maintain a catalogue of space objects. In this context, survey radars can provide angular track, slant range, and Doppler shift measurements without the need for transit prediction, allowing either the refinement or the initial determination of the target orbital state. In the latter case, a proper Initial Orbit Determination (IOD) technique is required to reconstruct the orbital state of the observed object. This work presents the IODAD algorithm (Initial Orbit Determination from Angular and Doppler shift measurements), a novel radar IOD method when slant range is not available, and thus relying only on the angular and Doppler shift measurements. The proposed IOD algorithm combines the optical admissible region, computed from the angular track measurement, with the measured Doppler shift to compute a first estimate of the orbital state. This combination forks depending on whether the radar is monostatic or bistatic. At the end, the first estimate is refined through a batch filter and the IOD result is returned in terms of mean state and covariance. Unlike existing methods, the new algorithm offers greater flexibility and ease of operational application, as it does not need long measurements tracks as input, nor a specific advanced computational technique. Numerical simulations show the potential of the IODAD algorithm, both through nominal and sensitivity analysis, proving its validity to any survey radar. In addition, a comparison with an existing method demonstrates the significantly better performance of the proposed method. Finally, the results are confirmed by analysing a real dataset of transits concerning calibrator satellites.

The increasing population of resident space objects is currently fostering many space surveillance initiatives. In this framework, on-ground multireceiver radars allow to reconstruct the target angular track, but the array configuration may cause the presence of multiple solutions and, if no pass prediction is available, the ambiguity cannot be solved a-priori. This work proposes an evolution of the Music Approach for Track Estimate and Refinement (MATER) algorithm. Given two different signals reflected by the same target, at each observation epoch their Direction Of Arrival (DOA) is estimated from the signal Covariance Matrix (CM) through the MUltiple SIgnal Classification (MUSIC) algorithm. Then, the possible ambiguous estimations are solved through the delta-k technique: the correct DOA is considered as the one featuring the smallest angular deviation comparing the two CM results. This process is repeated for all the epochs, and the DOAs are clustered according to the RANdom SAmple Consensus (RANSAC) algorithm. Finally, the most populated cluster is considered as the correct one, and the angular track is computed through a time regression of the two angular coordinates. The evolution of MATER algorithm is tested through numerical simulations. The algorithm converges to the correct solution in 100% of the cases, with an angular accuracy in the order of 1–10 mdeg.