The Signal Detection And Sortation Technology Research Of Passive Interception Equipment

Electronic warfare, as the fifth warfare after land battle, sea warfare, space warfare and air warfare, is currently an important representative of a country's military strength. Electronic reconnaissance is the basis to won electronic warfare, while signal detection and sorting are the main elements for electronic reconnaissance system which is the basis to won electronic warfare, so it is the key points for successful electronic warfare to detect signal accurately and to do signal sorting precisely. In recent years, artificial intelligence algorithms such as neural network, genetic algorithm emerge in endlessly, but because of the limitation of real-time in engineering practice, some traditional classic algorithms still occupie the important position.This article aims to detect signal after the data channelized that received by actual passive interception equiment. Because the autocorrelation detection method is strong real-time performance, easy to engineering practice, so choose the autocorrelation detection method to detect signal. For autocorrelation detection method can not adapt multichannel signal detection as it is not applicable, this paper use N-point adaptive sectional autocorrelation detection; Meanwhile, improved autocorrelation detection method has higher detection probability than sliding window energy accumulation detection method under the condition of certain SNR, but the autocorrelation detection method has bigger detection error if the two methods can detect signal at the same time. Therefore, the detection method that combine autocorrelation detection method and sliding window energy accumulation detection method is adopted to detect, detection probability is improved and detection error is reduced.Besides, the combined method has strong timeliness and is easy to engineering practice.According to parameter measurement after envelope combined, finally worked out the Pulse Description Words (PDW) in order to pre-sorting and sorting.Signal sorting consists of pre-sorting and sorting. According to the PDW received by signal detection, the carrier frequency and pulse width are simultaneously pre-sorted by K-Means algorithm. The two defects that initial clustering center and clusters number of K-Means cannot be determined and abnormal data (isolated point and noise point) effect algorithm shall be improved-abnormal data are removed through grid method, next cluster center and cluster number can be initialized through bilayer difference algorithm, the times of iterate in the process of K-Means algorithm is reduced and the accuracy of cluster results is improved. Finally signal sorting can be done to presort results by SDIF algorithm. Radiation source signal of some complex operation modes like frequency agility, PRI staggering/jittering can be accurately sorted by simulation analysis and verification.This paper uses experimental data and actual received data to simulation and comparative analysis the signal detection and sorting algorithm, first using the detection method that combine improved autocorrelation detection method and sliding window energy accumulation detection method to detect the actual data, extraction envelope, work out the PDW, and then using improved K-Means algorithm to pre-sort, using SDIF algorithm to sort.Finally demonstrate the effectiveness of the algorithm fully after these operations.