Track-to-track association methodology for operational surveillance scenarios with radar observations

This paper proposes a novel track-to-track association methodology able to detect and catalogue resident space objects (RSOs) from associations of uncorrelated tracks (UCTs) obtained by radar survey sensors. It is a multi-target multi-sensor algorithm approach able to associate data from surveillance sensors to detect and catalogue objects. The association methodology contains a series of steps, each of which reduces the complexity of the combinational problem. The main focus are real operational environments, in which brute-force approaches are computationally unaffordable. The hypotheses are scored in the measurement space by evaluating a figure of merit based on the residuals of the observations. This allows us to filter out most of the false hypotheses that would be present in brute-force approaches, as well as to distinguish between true and false hypotheses. The suitability of the proposed track-to-track association has been assessed with a simulated scenario representative of a real operational environment, corresponding to 2 weeks of radar survey data obtained by a single survey radar. The distribution and evolution of the hypotheses along the association process is analysed and typical association performance metrics are included. Most of the RSOs are detected and catalogued and only one false positive is obtained. Besides, the rate of false positives is kept low, most of them corresponding to particular cases or objects with high eccentricity or limited observability.