Reseach On Adaptive Waveform Design And Array Optimization For MIMO Radar

Multiple input multiple output(MIMO)radar has become an advancement in the field of radar signal processing,which attracts the attention of researchers in recent decades due to its waveform diversity and flexible array configuration.With the development of MIMO radar,the orthogonal waveform is not the only choice.Nowadays,the probing waveform can be designed by actual needs,such as maximizing the power around the interested locations,or approximating a desired beampattern.Two beampattern design approaches are available in the literature: beampattern matching and minimum sidelobe design.For the beampattern matching,the high sidelobe level problem is not well considered.Furthermore,this method focuses on the design of beampattern but ignores the adaptive processing requirements,which limits its application such as the range sidelobe suppression in the time domain and the anti-active jamming in the frequency domain.Subarray and sparse array are two common used array configuration in MIMO radar.The subarray based Phased-MIMO radar only focus on the improvement of SNR(signal to noise ratio),but ignore the anti-interference ability.The sparse array based MIMO radar is able to obtain the additional degrees of freedom.However,the optimization of sparse array is too difficult to be addressed by the existing algorithms due to the complexity of beampattern design.This thesis focuses on the study of adaptive waveform design and array optimization for MIMO radar and proposes several new solutions for the existing problems.Firstly,two algorithms including Iterative Diagonal Approximation Algorithm(IDAA)and Arbitrary-Dimensional Iterative Spectral Approximation Algorithm(ADISAA)are proposed to address the high sidelobe level problem and the optimization for the beampattern matching and the range sidelobe masking,respectively.In the IDAA,the beampattern can be modeled as a convolution of the desired beampattern and the reciprocal steering vector matrix.Based on this model,a matching beampattern can be established to approximate the desired transmit beampattern using the iterative diagonal approximation.In the ADISAA,Multi-WECAN(Multi-sequence weighted cyclic algorithm-new)is derived as a projection operator to design a good correlation properties at specific lag for suppressing the range sidelobe besides IDAA.Thus,the optimization of multi-function narrow band waveform design can be achieved.The combination of the IDAA and ADISAA provides an efficient waveform design framework for the beampattern matching and the adaptive constraints optimization,which is more suitable for adaptive waveform design in MIMO radar.The simulation results present that the ADISAA based beampattern can sacrifice a small portion of beampattern in exchange for deeper notch of correlation.Besides,the sidelobe notch can be more ideal and flexible.Secondly,an ADISAA framework based multi-function stopband waveform method is proposed to address the beampattern matching,range sidelobe and active jamming problems.The use of this novel method provides a matching beampattern as well as good correlation properties at specific lag and power spectrum notch for stop band transmitting waveform.Taking into account of the two-dimensional space broadband waveform caused high sidelobe level problem,in this thesis,the frequency and bandwidth are not strictly restricted but considered as a projection operator variable by deriving the rank deficient Fourier transform.This proposed waveform design method is considered as a part of ADISAA to be combined with the other two projection operators including beampattern and good correlation properties at specific lag to achieve the optimized multi-function stopband waveform design.Thirdly,a new method for polarized beamforming is proposed to address the suppressing interference,then a new method based on the iterative matrix spectral approximation algorithm(IMSAA)is used to solve frequency diversity waveform optimization problems.The proposed polarized beamforming imposed orthogonal polarization constraints on the null,which can effectively improve the anti-interference ability for Phased-MIMO radar.The frequency diversity signal among subarray unit is optimized by the proposed IMSAA to generate an orthogonal signal with good correction.Numerical simulations have demonstrated the effectiveness of the proposed algorithms.Finally,a novel approach based on multi-objective differential evolution algorithm(MODE)is proposed to address the paralled optimization of beampattern matching and peak side-lobe suppression for the sparse array design in MIMO radar.In the first stage,the cyclic algorithm(CA)is used to obtain a covariance matrix that satisfies the beampattern approximation for a full array.In the second stage,based on the covariance matrix obtained in the first step,the single-objective optimization of the sparse antenna array is carried out to synthesis the beampattern by using the genetic algorithm(GA)and differential evolution(DE)algorithm.Since the peak side lobe level(PSLL)is unable to be suppressed by the former algorithms,this paper consider the PSLL suppression as an inequality constraint to be a multi-objective optimization problem.To address this new problem,the MODE is proposed.Simulation results show that beampattern matching and PSLL suppression can be optimized in the parallel implementation using MODE.