Detection And Identification Of Frequency Hopping Signal

Frequency-hopping communication has the advantages of strong anti-fading and anti-interference ability,low interception probability,high confidentiality,and secure networking capability,etc.It is widely used in military and civilian fields.However,these advantages of frequency-hopping communication also bring difficulties to intercept and decipher for non-cooperative reception.Aiming at the problem of non-cooperative reception of frequency-hopping signals,this paper studies the methods of frequency-hopping signal detection,parameter estimation,and frequency-hopping station selection.The specific research work and results are as follows:1.A method of frequency hopping signal detection based on interference elimination is proposed.For the detection of frequency hopping signals in a complex electromagnetic environment,the interference signal is first removed,and then the frequency hopping signal is further identified using the time-frequency characteristic curve method based on the difference in time-frequency characteristics of the frequency-hopping signal and other modulated signals.Simulation results show that when the SNR is 2d B,the recognition rate of the proposed method can reach 100%,which is better than the detection algorithm based on time-frequency analysis by 2d B.2.A parameter estimation method based on the decimation factor for time-frequency ridge stitching is proposed.For the slow frequency hopping signals,we used the non-overlapping segmentation method to segment the signal,and then perform a spectral conversion based on the decimation factor for each segment.And the time-frequency ridges obtained from each section are stitched together for parameter estimation,thereby reducing the algorithm's Computation.Finally,the influence of extraction factors on parameter estimation is analyzed.The simulation and measured data show that under the same decimation factor,compared with the maximum stitching method,the time-frequency ridge stitching method has stronger anti-noise ability,improves the parameter estimation performance,and reduces the complexity of the algorithm.Besides,for the missing frequency hopping points,the exact frequency set is obtained by correcting the disordered frequency hopping points.Simulation results show that the method can accurately estimate the frequency set when the SNR is higher than-5d B when there is only hop loss.3.A parameter estimation method based on Fractional Wavelet Transform(FRWT)is proposed.Firstly,the time domain signal is denoised by the FRWT decomposition method.Then Hilbert-Huang transform(HHT)is used to perform Hilbert spectrum analysis on the denoised frequency-hopping signal to extract the parameter features.This method solves the problem that the time-frequency resolution of the traditional time-frequency transform method is not high when the adjacent frequency interval of the frequency hopping sequence is relatively small.The simulation results show that the SNR of the FRWT process is about3d B lower than that of the Discrete Wavelet Transform(DWT)method when the relative error value is 10^-2.4.A method of frequency hopping station identification based on contour features is proposed.First,extract the contour features of the signal on the contour map,then construct and preprocess the contour map based on the contour features,and finally input the contour map to the convolutional neural network(CNN)for training and testing and then realizes classification recognition.This method solves the problem that the existing frequency-hopping station sorting method has the complexity of manually extracting parameter features.Simulation results show that when the sorting rate is 100%,its signal-to-noise ratio is-15d B after cropping,which is 10d B lower than both support vector machine(SVM)and traditional K-Means clustering algorithm.The algorithm verification of the measured data shows that the method can correctly classify the four types of drones:Phantom 4 Pro,hm UAV,SYMA?X8HW,and Inspire 2.