New Methods Of Radar Weak Target Signal Detection And Their Applications

The most basic purpose of radar is to detect a target or phenomenon.However,in the modern battlefield,Radar Cross Section(RCS)of weak target,such as stealth aircraft,anti-radiation missile and unmanned aerial vehicle,is very small,which results in low signal to noise ratio(SNR)of target echo signal and makes the target difficult to be detected.Generally,increasing integration time(long time coherent integration of single frame and non-coherent integration of multiframe)is an effective approach to improve SNR of echo and detection performance.However,in long time coherent integration of single frame,the range migration(RM)and Doppler frequency migration(DFM)may occur for high speed and maneuvering targets,which can make the traditional coherent integration algorithm invalid.Although there are no RM and DFM in non-coherent integration of multiframe,this method has to face many other challenges,such as heavy computational burden and complex threshold determination.Therefore,how to improve the detection performance of radar to weak target signal is the main purpose of this thesis.This thesis mainly focuses on discussing the existing problems in long time coherent integration of single frame and non-coherent integration of multiframe.The main reseaches and innovation are listed as follows:1)To compensate the RM and DFM of target with constant acceleration,a novel coherent integration method,i.e.,keystone transform-conjugate time reversing transform(KTCTRT),is proposed.Firstly,the proposed algorithm employs the keystone transform(KT)to remove the RM efficiently.Subsequently,the conjugate time reversing transform(CTRT)is adopted to remove the DFM.Finally,the Fourier Transform(FT)is conducted to achieve the coherent integration.The proposed method doesn't require parameter searching and its computation is in the same order with the KT.Simulations and real data confirm the effectiveness of the proposed algorithm.2)To compensate the RM and DFM of target with variable acceleration,a novel coherent integration method,i.e.,short-time fourier transform-cubic phase function(STFTCPF),is proposed.Firstly,the proposed method employs short-time fourier transform(STFT)to construct a special compensation function.With the compensation function,the range migration(RM)can be compensated.Then the cubic phase function(CPF)is applied to estimate the target's acceleration and jerk.With the estimated acceleration and jerk,the all DFM can be compensated.Finally,the coherent integration and target detection are performed.The STFT operation before the CPF can enhance the detection performance for the multiple targets in the same range cell and for the low SNR target.Moreover,the CPF involves only one-dimensional searching and further reduces the computational complexity.Simulations and real data confirm the effectiveness of the proposed algorithm.3)To avoid challenges about the computational burden and the complex threshold determination in dynamic programming algorithm based track-before-detect(DP-TBD).A novel greedy algorithm based track-before-detect(G-TBD)is proposed for weak target detection and tracking in radar systems.The proposed G-TBD is simpler and easier to implement than DP-TBD.Even though sometimes the traditional greedy algorithm can only find suboptimal solution,it will have the similar performance to dynamic programming algorithm under some restrictions and then improve detection and tracking performance of radar.Detailed analyses,including its feasibility,parameter settings and detection performance,are presented.Additionally,the G-TBD is firstly applied to single frequency network(SFN)based MIMO passive radar(SMPR)in this thesis.Simulations and real data confirm the effectiveness of the proposed method.