PSO Main Ridge Slice Of Fuzzy Search Function Method And Sorting Of The

Radar emitter signal sorting is the key technology of radar electronic countermeasure system, and the core content of radar electronic countermeasure information processing. Along with the intense confrontation of radar electronic warfare, new complex system radar constantly put into using and gradually occupies the leading position. The signal density of electromagnetic environment has reached up to million pulses per second, radar working frequency coverage expand to both ends constantly. In such dense, complicated and changeful signal environment, radar emitter signal sorting, especially for all kinds of new type of complex system radar emitter signal separation has become the serious bottlenecks of electronic reconnaissance signal processing, and become the key problem to restrict electronic warfare equipment to improve the performance. This paper focus on the subject, mainly aimed at unknown radar situation, from the quick searching of ambiguity function main ridge slice for radar emitter signal, deeply discusses and researches the related theory of complicated system radar emitter signal sorting from feature extraction and performance analysis of AFMR slice. The main work and research results are as follows:(1) Firstly, this paper briefly introduces the research situation and challenges faced with at present of radar emitter signals sorting at home and abroad. Then, this paper briefly introduces the main several algorithm of radar emitter signal deinterleaving, that are the separation principle and the advantages and disadvantages of PRI sorting algorithm, multiple parameters associated sorting algorithm, clustering sorting algorithm and based on in-pulse characteristics sorting algorithm.(2) This paper briefly introduces the definition and properties of fuzzy function, the standard Particle Swarm Optimization algorithm and several common improved PSO algorithm, detailed introduces the extraction principle and process of AFMR slice for radar emitter signals, and constructs the improved PSO algorithm based on uniform initialization strategy, random inertia weight and natural selection to search the AFMR slice features quickly. The proposed method in significantly shortened the time of extracting AFMR slice features, still can obtain more accurate AFMR slice at the same time, and also has good stability and reliability in dynamic SNR environments, greatly improving the possibility of AFMR slice features as the effective supplementary of five radar emitter signal sorting classic parameters.(3) In this paper, using the local difference method to extract the three characteristic parameters of AFMR slice that searched fast by the improved PSO algorithm, that are the sum of difference sum_diff, the maximum of difference maxdiff and the distribution entropy of difference entr_diff, from local aspects represent the signal fuzzy energy distribution difference information. Then using Fuzzy-C Mean cluster algorithm to do clustering performance analysis about the extracted characteristic parameters. Matlab simulation results show that the extracted features not only have strong compactness within clutters and large separation between clutters when the SNR is not lower than0db, but also have good performance of anti-noise. These local difference corresponding characteristic parameters are likely to become the effective complement of five classic parameters for radar signal sorting, and also can be used as the references for later radar emitter signals sorting and identification.