Research And Implementation Of UAV Detection Technology Based On Signal Features

With the progress of science and technology,the civilian UAV has been greatly popularized and developed,applied in many fields of production and life.At the same time,the intentional or unintentional illegal flight behavior of UAV has a serious impact on social and national security.Aiming at the problem of monitoring the illegal flight behavior of UAV,this thesis studies the UAV detection technology based on the characteristics of radio signal.The specific research results are as follows:1.A UAV detection and recognition method based on multi-feature matching of flight control signal is proposed.For the detection and recognition of flight control signals in a complex electromagnetic interference environment,the three characteristics of the flight control signal’s bandwidth,duration time point,and frequency of occurrence are extracted,and identified by matching with the feature database.Specifically,the time-frequency analysis is carried out by short-time Fourier transform.The ridge line range value and ridge line frequency value of each segment are calculated by signal preprocessing methods such as noise removal,spectrum reconstruction,anti-aliasing and ridge line extraction.Then,the frequency difference method combined with the bandwidth feature is used to find the signal duration time point characteristics,then the frequency characteristics of the signals are obtained by the spectral graph statistics method,and the above features are matched with the database to complete multi-objective recognition in the library,while for the UAVs not in the library,the custom judgment criteria are used to detect.The experimental results show that all the test targets can meet 100% recognition rate when the SNR is-15 d B.2.A UAV detection and recognition method based on the two-stage structure of image transmission signal is proposed.The first stage is the blind detection algorithm for suspected image transmission signals,which uses segmented Fourier transform,spectrum reconstruction,spectrum integration,signal threshold judgment,leak point completion,bandwidth judgment and other methods to identify the suspected UAV image transmission signal.The test results show that the detection rate is 100% when the SNR is-22 d B.The second stage is the accurate recognition algorithm of image transmission signal,which is used to distinguish small features on the basis of the first stage.According to the periodic characteristics of UAV image transmission signal,the sliding shift cyclic autocorrelation is used to extract the corresponding features,and then the matching recognition is carried out.The experimental results show that the recognition rate of UAV can reach 100% when the SNR is-4d B.3.A method for UAV individual identification based on Radio Frequency Distinct Native Attribute(RF-DNA)is proposed.This method extracts the transient part of the signal by receiving the flight control and image transmission signals of the UAV,using signal variance trajectory change point detection,fractal Bayesian change point detection,and phase variance trajectory change point detection.Then,the RF-DNA feature statistics is carried out and the feature database is established.The features include the standard deviation,variance,skewness and kurtosis of instantaneous amplitude,instantaneous frequency and instantaneous phase.Then the feature dimensionality reduction is performed by principal component analysis(PCA).Finally,the multi-classification support vector algorithm(SVM)is used for classification and identification.The experimental results show that the recognition rate of this method can reach 97% when the SNR is 20 d B for different types of UAVs,and 98% when the SNR is 20 d B for the same type of UAVs.4.Designed and implemented an FPGA-based passive detection system for UAVs.The systems are radio signal acquisition subsystem,flight control signal rapid detection and identification subsystem,and dual-source parallel detection and identification subsystem.For the radio signal acquisition subsystem,K7 series FPGA,AD9361,DDR3,Gigabit Ethernet,and custom serial ports are used to form an acquisition card based on a zero-IF receiver.The test results show that the acquisition card can meet the requirements of 70MHz-6GHz radio Signal acquisition,storage,and transmission.The flight control signal rapid detection and identification subsystem combines the flight control identification algorithm with the radio frequency spectrum acquisition subsystem,and uses FPGA as the core processor to implement the algorithm.For the dual source parallel detection and recognition subsystem,the top-down architecture and multi-channel parallel algorithm processing mode are adopted for the parallel detection of flight control and image transmission signals respectively.The outfield test of the UAV passive detection system are completed.The results show that the system can quickly and accurately identify multiple UAV flight targets in the protection area by receiving and processing UAV flight control and image transmission signals.