Automatic pulse repetition pattern reconstruction of conventional radars

Abstract

Signals of radars with fixed, stagger and other conventional pulse repetition intervals (PRI) occupy a large proportion in most radar reconnaissance datasets. Automatic and accurate reconstruction of the PRI patterns of these radars is an important task in realizing intelligent processing of radar reconnaissance data. In order to address this problem, this article proposes an automatic method based on the extraction and extension of frequent differential time-of-arrival (DTOA) items, which can be combined with other pulse parameters such as pulse width. The values and ordering of the frequent items are determined to reconstruct a complete PRI cycle, and the radar pulses are deinterleaved at the same time as a by-product. Compared with the commonly used one-dimensional radar reconnaissance parameters such as framework PRI or typical PRI set, the reconstructed PRI pattern contains more information about radar operational modes and can be applied to realize online and parallel pulse deinterleaving and more accurate radar type classification.