Quasi-optimal adaptive algorithm for filtering the coordinates of a moving source of radio emission during single-position passive direction finding on sea over-the-horizon paths

Annotation: The work provides an experimental substantiation of the statistical model of passive bearing only measurements in conditions of receiving UHF signals from a mobile source of radio emission on over-the-horizon sea paths with a length of up to 300–500 km. It is shown that the sequence of bearing readings to a source operating in the scanning mode of the antenna pattern is unequally accurate. It is accidentally «clogged» with anomalous errors. For the proposed model, based on the separation principle known in the Markov theory of nonlinear filtering, a multi-channel algorithm for estimating the current coordinates and parameters of the radiation source’s movement is presented, that is adaptive to the unknown speed and heading of the object’s movement. The performance of the algorithm is shown taking into account the random contamination of current bearing estimates with anomalous errors in the presence of random omissions, and estimates for the relative mean square error of the current range are obtained.

Keywords: estimation of coordinates and motion parameters, adaptive algorithm, anomalous errors, coordinate estimation, over-the-horizon reception, passive direction finder