In the context of Space Surveillance and Tracking the observations provided by a network of sensors are processed and used to maintain a catalogue of orbital elements. The build-up of a catalogue and its maintenance depend on the capacity to determine the orbit of the observed objects from few measurements sparsely distributed on relative long arcs. The sparse observations or short sequences of observations need to be correlated or associated with each other. The association consists in identifying the series of observations, so-called tracklets, that belong to the same space object.
A method is proposed to associate radar tracklets, consisting of radar observations, to other radar tracklets or to optical tracklets acquired with optical sensors and consisting of angles-only observations. The association method is applied to different simulated observation scenarios and the performance of the algorithm is evaluated.