Research On Rapid Algorithm For Sorting And Identification Of Unknown Radar Signal

Radar signal sorting refers to distinction of all received signals from different radar emitters and parameter estimation, detection and identification of signals belonging to the same radar emitter after sorting. Thus information supporting for interference of radiation sources and anti-radiation attacks is provided. Obviously it is not only a key part in signal processing of radar reconnaissance, interference guidance and anti-radiation guidance, but also a bottleneck problem of radar reconnaissance processing in a complex electromagnetic environment. So radar signal sorting plays a important role in seizing control power of electromagnetism and information for the battle of modern high-tech warfare and information warfare.Radar signal sorting based on dectectable characteristics of radar signals, particularily which holding both special nature of certain source and beneficial to using modern electronic measurement techniques for fast, accurate, high-resolution measurement. It is generally characterised by angle of arrival (AOA), radio frequency (RF), pulse width (PW), time of arrival (TOA), pulse amplitude (PA), pulse repetition inerval (PRI), pulse modulation characteristics (PIM), antenna scanning feature (AS) . PRI and AS based on the measurement of TOA and PA are obtained by data secondary processing and PIM is obtained by time-frequency analysis of a signal waveform feature. In order to improve processing speed of radar detection signal, pre-sorting for the characeristics of AOA, PF, PW obtained by direct measurement is performed. Then deinterleaving of radar pulse repetition interval and radiation dectetion is performed.This thesis concentrates on fast algorithm for identifing and estimating pluse repetition interval (PRI) of unknown radar signals. Firstly, radar signal environment model (a description of radar emitter ) and PRI model (single-and multi-radiation environment) have been discussed. Secondly, efficient deinterleaving algorithms for different radar emitters in the complex environment have been analysed and simulation algorithms of sequence search, traditional histogram, CDIF and SDIF, PRI transform have been accomplished. The relationship between correct probability of deinterleaving and identifying radar pulses and radar environment has been statistically analysed. Finally, a comprehensive deinterleaving algorithm of pulse trains with improved detection threshold to improve the accuracy of PRI deinterleaving based on the advantages and disadvantages of various algorithms has been propsed.