Font Size: a A A

Research On Key Technologies In Multi-Level Processing For Electronic Reconnaissance Of Radar

Posted on:2016-08-13Degree:DoctorType:Dissertation
Country:ChinaCandidate:C LiFull Text:PDF
GTID:1108330509961067Subject:Information and Communication Engineering
Abstract/Summary:PDF Full Text Request
As an important part of radar electronic countermeasure, radar electronic reconnaissance plays a significant role in modern electronic warfare(EW). With the development of radar technology, new system radars, such as phased array radar and netted radar, have been widely applied in modern EW and brought new challenge to current radar electronic reconnaissance. Tranditional technologies of radar electronic reconnaissance, such as radar signal sorting and identification, mainly focus on the processing of the radar signal. Although these techniques are still the base of electronic reconnaissance of all kinds of radar system, they are no longer meet the demand of electronic reconnaissance of modern new system radars. Therefore, based on the background of the electronic reconnaissance of new system radars, such as phased array radar and netted radar, we research the key technologies of radar electronic reconnaissance in signal layer, data layer and information layer respectively in this thesis.In the aspect of the processing in signal layer, we mainly treat the sampled signal in intermediate frequency and study the problem of pulse detection and parameters estimation of radar signal, thus accomplishing the processing from signal to data. First of all, the edge detection principle in image processing is applied to the pulse detection problem in digital signal processing, and a pulse detection metheod based on difference of boxes filter is presented. This method is simple and can be applied to the pulse signal of various types. In the estimation of signal parameters in time domain, such as time of arrival(TOA) and pulse width(PW), the estimation precision by this method is higher than that by traditional energy detection method. Then, based on the analysis of typical linear and nonlinear time-frequency analysis methods, we propose an instantaneous frequency(IF) estimation method of radar signal by combining the Wigner-Wille distribution(WVD) and the generalized S transform(GST). This method not only keeps the high time-frequency resolution of the WVD, but also takes the advantage of the GST of suppressing the cross terms. Hence, robust and high precision IF estimation results can be obtained.In the aspect of the processing in data layer, we mainly treat the parameters of radar pulse signal and study the problem of radar antenna scan type(AST) recognition and radar words extraction of multi-function radar, thus accomplishing the processing from data to information. Firstly, we construct a simulator of radar AST. Then, by means of the time-amplitude parameters of radar pulse signal, we present methods of automatic recognition of electronic scan and mechanical scan respectively. So that the one-dimensional electronic scan, the two-dimensional electronic scan and 8 kinds of mechanical scan can be identified, and the parameters of one-dimensional electronic scan can be estimated. Finally, we study the radar words extration method by means of the pulse description words(PDW). The method realizes the extraction of radar word in data layer, and provides a basis for syntactic pattern recognition of multi-function radar.In the aspect of the processing in information layer, we mainly treat the obtained electronic reconnaissance information and study the problem of radar netting type recognition and association rules mining of reconnaissance information, thus accomplishing the processing from existing information to new information. At first, we put forward a sequential recognition method of radar netting type based on the fusion of radar information and communication information. Through the selection of main radar information and communication information, we extract the relative characteristic parameters and calculate the matching degree of the feature vectors. Subsequently sequential recognition of radar netting type is implemented based on D-S evidence information fusion. Then, we study the application problem of association rules mining in radar electronic reconnaissance information processing. According to the characteristics of radar electronic reconnaissance, we preprocess the reconnaissance information, and do some adjustments to classical algorithm of association rules mining- Apriori algorithm. At last, the results of the association rules mining are analyzed.This thesis studies the multi-level processing technologies in radar electronic reconnaissance respectively in signal layer, data layer and information layer, which enriched the research system of radar electronic reconnaissance. It also has important practical significance for the electronic reconnaissance of new system radars, especially for phased array radar and netted radar.
Keywords/Search Tags:Phased Array Radar, Netted Radar, Pulse Detection, Instantaneous Frequency Estimation, Antenna Scan Type, Extracting of Radar Words, Netting Type Recognition, Association Rules Mining
PDF Full Text Request
Related items