Research On Radar Signal Sorting Algorithms Under Dense Signal Environment

Radar signal sorting is a key technology of radar electronic surveillance system,system,continuously optimizing signal sorting method helps to improve the signal processing abilityof reconnaissance system. Some measurable characteristic parameters carried by radar signalcould be considered as the basis of radar signal sorting, which are mainly pulse time ofarrival(TOA), angle of arrival(AOA), pulse width(PW), radio frequency(RF), pulseamplitude(PA) and so on.The traditional method of radar signal sorting is to deal with TOA toget the pulse repetition interval(PRI), so as to carry out the signal separation according to thePRI feature. In order to enhance the signal processing speed and sorting accuracy, AOA, PWand RF are chosen as the pre-sorting parameters,and then to implement the main sorting withPRI.This paper mainly studies the algorithms of radar signal sorting under dense signalenvironment, including introducing the above signal sorting parameters, and describing howto realize the digital simulation of radar signal environment. On the basis of simulationmodeling, several kinds of traditional methods based on PRI are compared and analyzed. Withthe purpose of adapting to the demand of modern battlefield environment, many scholars havebeen exploring and researching on realtime and effective method of radar signal sorting,including improving the traditional algorithms, using multiple parameter to achieve the signalsorting, adopting blind source separation technology and pulse feature extraction and so on.Here the paper researches the signal sorting method based on multiple parameter.So severalcommon signal sorting methods based on multiple parameter are introduced, on account ofthese algorithms for signal sorting are in a passive state, that is to say these algorithms need toset a series of principles to determine whether data objects belong to the same class, aclustering method based on set pair analysis is put forward for radar signal sorting. Set pairanalysis method study the subject directly from the similarity between data objects, so it hasthe advantage of fast calculation speed. In consideration of the fact that the clustering resultsof original clustering method based on set pair analysis for radar signal sorting is unstable, sothe paper improves this method, and verifies the method through simulation. For purpose offurther study on the performance of the improved method, the paper also studies the improvedmethod under the condition of different noise rate.In addition, when using support vector clustering algorithm for radar signal sorting, inorder to avoid some disadvantages, this paper describes a improved algorithm based onsupport vector clustering for radar signal sorting, which combines support vector clusteringwith improved set pair analysis method. Not only does this algorithm effectively avoid severaladverse factors, but also improve the anti-noise capability of the improved set pair analysismethod, and the simulation verifies the algorithm.