Study On High-Speed Low-Observable Target Detection

High-speed weak target has the typical characteristics of supersonic speed,strong maneuverability,and low radar cross section,which bring about severe challenges for the current radar detection technique.Considering the coverage,survivability,and development costs of the radar,long-time coherent integration methods are adopted to improve the detection performance of the high-speed low-observable target.However,the range migration will occur due to the hypersonic characteristic.The strong maneuverability may also lead to a degree of range migrations and serious Doppler frequency migrations.Because of the range migrations and Doppler frequency migrations,the detection ability of the radar will degrade.This dissertation studies on the detection algorithms of the high-speed low-observable target under different motion states.The study contents of this dissertation are summarized as follows:1.With respect to the serious range migration under the uniform motion model,a novel detection algorithm for high-speed low-observable target based on frequency-domain realized modified axis rotation moving target detection(MAR-MTD)is proposed.MAR-MTD is the modified method based on the axis rotation moving target detection(AR-MTD)and improved axis rotation moving target detection(IAR-MTD)methods to solve rotation errors induced by the coordinate transformations.Different from AR-MTD and IAR-MTD,MAR-MTD selects the image center of 2-D compressed signal as the origin to construct the coordinate system,and changes the axis rotation matrix to eliminate the coupling between the fast-time axis and the rotation angle.In order to further reduce the rotation error,the frequency-domain implementation of the MAR-MTD is also given.Compared with the AR-MTD and IAR-MTD,MAR-MTD with frequency domain implementation has a better detection performance.2.For the range migration and Doppler frequency migration under the accelerated motion model,a new method based on Keystone transform and linear canonical transform(KT-LCT)is proposed.The KT is first utilized to correct the linear range migration(LRM),and then the LCT is exploited to compensate the linear Doppler frequency migration(LDFM).With respect to the velocity ambiguity phenomenon,an improved method combining the coarse search and fine search is presented.The proposed algorithm can be extended to multi-target detection scenes.The theoretical analysis and simulations prove that the proposed method has the ability to detect the high-speed maneuvering target in low signal-to-noise(SNR)conditions.Compared with the second-order Radon-Fourier transform(So RFT)and Radon Fractional Fourier transform(RFRFT),the proposed method can effectively reduce the computational load under the guarantee of the close detection performance.3.For the LRM,undersampled linear Doppler frequency migration(ULDFM),quadratic Doppler frequency migration(QDFM),and the velocity ambiguity phenomenon faced with the ground-based radar detection under the jerk motion model,a new detection algorithm for high-speed maneuvering low-observable target based on the fast discrete chirp-Fourier transform(FDCFT)is proposed.To correct the LRM,the KT is first utilized to remove the couple between the range frequency and slow time.Thereafter,the FDCFT is proposed to compensate the ULDFM and QDFM simultaneously.Based on the concept of decimation-in-time fast Fourier transform(FFT),the FDCFT can reduce the computational load by performing the time-domain radix-4 decomposition along the searching chirp rates and derivative of chirp rates.The proposed algorithm also gives the detailed analysis of the half blind velocity effect and the incomplete correction of range migration induced by the velocity ambiguity phenomenon.Furthermore,the compensation functions are constructed to solve the above problems.Compared with the third-order Radon-Fourier transform(To RFT),the proposed method can achieve a close detection performance but with relatively low computational costs.Moreover,the FDCFT is also extended to the compensation of arbitrary higher-order Doppler frequency migration.Theoretical analysis indicates that the savings achieved through the extended FDCFT are more significant as the order of motion model increases.4.A novel method based on efficient angular chirp-Fourier transform(EACFT)is proposed to correct the complex range migrations and Doppler frequency migrations caused by the jerk motion model under the condition of high-resolution detection scenario.The angular chirp-Fourier transform(ACFT)is first derived from the chirp-Fourier transform(CFT)and the fractional Fourier transform(FRFT).Thereafter,the EACFT is proposed by utilizing the time-frequency rotation characteristic.The EACFT can realize the detection and parameter estimation of linear frequency modulation signal through a few 1-D searches,which avoids the traditional 2-D search.To improve the applicability of the EACFT in the low SNR and multi-component circumstances,the MF-EACFT and SS-EACFT are respectively proposed by combing the EACFT with the morphology filtering(MF)and spectrum smoothing(SS)techniques.Moreover,the combined method SS-MF-EACFT can own the respective properties of the MF-EACFT and SS-EACFT.Finally,the EACFT is applied to the high-speed maneuvering detection,and a new detection algorithm is proposed,where the adjacent cross correlation function(ACCF)is first utilized to remove all the range migrations and QDFM.The compensation function is then constructed to determine the adjacent cross-correlation peak.Finally,the SS-MF-EACFT is exploited to compensate the residual Doppler frequency migration.The effectiveness of the proposed algorithm is verified by simulated experiments.