| Radar sorting is an important link in modern electronic warfare.With the rapid development of electronic technology,the emergence of new radars and the continuous progress of radar technology,the electromagnetic environment of electronic warfare becomes more and more complex.How to sort the interlaced radar pulses correctly and quickly in the complex electromagnetic environment is still a big problem in radar sorting at present.In this paper,radar signal sorting in different scenes is studied.Firstly,the radar sorting algorithm in simple scenes is introduced,and a sorting algorithm based on time vector and sliding window is given.The accuracy and real-time performance of this algorithm are greatly improved.Then,a sorting algorithm based on hierarchical density clustering and relative spectral gap is proposed for the scene with high pulse density and complex modulation mode of radiation source.Finally,the simulation experiments of the two algorithms are carried out,and the experimental results show that the proposed algorithm can realize signal sorting well.This paper includes the following contents:(1)Aiming at the problems of traditional histogram algorithm in sorting dithered sequences,such as poor effect,large amount of calculation and difficulty in setting parameters,a sorting algorithm based on time vector and sliding window is given.The algorithm calculates and adjusts the time vector,and keeps the time difference with the highest frequency in the time vector.According to the relative position of the time difference in the adjusted time vector,the related pulses before and after can be quickly matched,thus reducing the complexity of the algorithm.The receiver can receive a large number of pulse signals in a short time,and divide the pulse stream into different sliding windows.The correlation between the pulses in the front and back sliding windows is strong,so the data in the next sliding window can be sorted by weight according to the sorting result of the previous sliding window.Experimental results show that this algorithm is better than the traditional histogram algorithm in stability and real-time.(2)In view of the poor real-time performance of single-parameter sorting algorithm in sorting dense pulse streams,a multi-parameter sorting algorithm based on hierarchical density clustering and relative spectral gap is given.Using pulse descriptor for clustering,the pulse stream is mapped to different cluster spaces quickly.The density information and relative existence time information of each cluster are deeply mined,and then the distance between clusters is redefined to obtain a adjacency matrix.Then,the number of radiation sources is determined by Laplace spectrum information of the adjacency matrix,and finally K-means clustering is performed to complete sorting.The experimental results show that the algorithm can complete the sorting task quickly and accurately in the scene of dense pulse flow,many chaotic pulses and many missing pulses.Compared with the algorithm proposed in(1)in different sorting scenarios,it is concluded that the single-parameter algorithm has better sorting effect on small data sets,while the multi-parameter algorithm has more advantages in real-time and stability when sorting large data sets.(3)A demonstration system of radar sorting algorithm is built based on Py QT5 GUI.The cumulative difference histogram,sequential difference histogram and the algorithm proposed in this paper are integrated into the demonstration system,and the feasibility of several sorting algorithms and the usability of the demonstration system are verified on the simulation data set.The analysis of simulation results further proves that selecting the appropriate sorting algorithm in different sorting scenarios can improve the sorting efficiency. |
