Study On Radar Signal Sorting Methods

Radar signal sorting is one of the most important steps to realize the deceptivejamming. With the development of modern electronic technology, the increasing densityand complexity of the radar signal bring big challenge for radar signal sorting. How topick up radar’s pulse serials quickly and exactly is a big problem facing by researchers.This paper mainly studies several radar signals sorting methods in the dense andcomplex environment. Though simulation in MATLAB, gain advantages anddisadvantages of each method. The methods studied in this paper state below:Firstly, this paper proposes an improved sequence difference histogram measurebased on PRI sorting methods. This paper studies the algorithms which includecumulative difference histogram(CDIF)、sequential difference histogram(SDIF) and PRItransform. Use these three methods stated above to sort the same the pulse strings anddiscuss advantages and disadvantages of them. Simulation outcome for the improvedSDIF sorting method shows that the improved SDIF can get better result than SDIF andfaster speed than PRI transform.Secondly, this paper studies the multivariant parameters cluster measures whichused for pre-sorting. Aiming at the defects of cluster methods which need someinformation previous to set tolerance, this paper proposes an adaptive tolerance radarsignals clustering method. This chapter discuss the multivariant parameters clustering、the K-means clustering and the SVM clustering. The above three methods are applied tosort same data. Based on the results, summarize advantages and disadvantages of eachmethod. Simulation outcome for the adaptive tolerance radar signals clustering methodshows that the method can also get good result without pre-knowledge.Thirdly, this paper combines above methods and proposes four sorting step flow.Discuss STFT and wavelet methods in time-frequency domain.Using the four step flowto sort emulational and real radar signals and get good results.