Radar Signal Sorting In Complex Electromagnetic Environment

With the continuous development of communication technology,the increase of electromagnetic equipment in space,the overlapping of signal bands and the increasing complexity of electromagnetic signals,it is an important topic to sort the received radar signals.In a complex electromagnetic environment,radar signals and other communication signals exist simultaneously in time and overlap each other in spectrum.In this paper,the mixed signals containing radar pulse signals are identified in the signal category,and the radar signals are jointly sorted.The main research content and work include the following contents:1.Separation and identification of mixed signals,pre-sorting for subsequent radar signal sorting.A mixed signal recognition algorithm based on time-frequency image and connected graph label is proposed,and the adaptive threshold is improved.The shorttime Fourier transform is used to transform the mixed signals.According to the different simultaneous frequency characteristics of different signals,the types of mixed signals are identified by time-frequency images and connected graphs.An improved adaptive threshold setting method is proposed to make it suitable for different cases of mixed signals,and radar signals are pre-sorted.Experiments show that for the mixed signals of three kinds of signals with equal amplitude,the accuracy of type identification is more than 90% when the signal-to-noise ratio is greater than 8d B.2.Sorting radar signals based on the improved PRI transform method.Aiming at the problem that the traditional PRI transform method is not effective in sorting the radar signal with jitter PRI value,two aspects of overlapping PRI boxes and update time starting point are improved.The PRI transform method needs to obtain the pulse arrival time in advance.A method to obtain the pulse arrival time series is proposed based on the actual received radar pulse signal.When the signal-to-noise ratio is greater than 0d B,the relative error of estimating PRI values for mixed radar pulse signals with different PRI values is about 9.5%.3.A method based on wavelet packet bispectral integration is proposed for radar signal sorting.The diagonal bispectral integral of the signal is calculated,and the characteristics of the modulation modes in different pulses are extracted by the bispectral integral of wavelet packet.The radar signals are sorted by the feature clustering.The experiment shows that the sorting accuracy can reach 100% when the signal-to-noise ratio of three different modulated radar pulse signals is greater than 4d B5.Two different radar features,PRI value and intra-pulse modulation type,were fused to classify the mixed radar signals are classified by support vector machine and cluster classifier.The experiment shows that the combination of the two features is better for the classification of mixed radar signals.When the signal-to-noise ratio is greater than0 d B,the accuracy of feature fusion sorting of SVM can reach more than 96%.