Target association with fuzzy inference on satellite electronic information

This paper proposes a target association algorithm with fuzzy inference on satellite electronic information, aiming at the association difficulties caused by the following problems: false alarms may exist in the observation, radar emitters can change radiating parameters and there are multiple platforms equipped with the same type of radar emitters in the observational scenario. Firstly, the electronic information is classified using the electromagnetic attributes and false targets are picked out to some extent by enforcing the spatiotemporal constraints. Secondly, according to the database of radar emitters, the classification results are recognized based on the radial basis function (RBF) neural network. Moreover, based on the similarity of the recognition results for different satellite electronic information, the initial association measurement can be constructed. Finally, to improve the association accuracy, an association modifying factor will be derived with fuzzy inference based on different kinds of appropriate auxiliary information. Experimental results on two groups of simulated data show its effectiveness and practicability.