Research On Automotive MIMO Radar Imaging Technology

The collocated multiple-input and multiple-output(MIMO)radar is widely been studied by many scholars from all over the world,with the excellent advantage of it,which has a broad application whether in the military or civilian field.The radar transmits a group of orthogonal waveforms to form a larger virtual aperture by the technique of orthogonal waveform separation in the receivers,and compared to traditional phased array radar,MIMO radar improves the angle resolution.At present,MIMO radar imaging suffers from low resolution,high computational complexity and low robustness and others.Recently,with the rapid development of millimeter-waveform radar in ADASs,many research corporates and institutions begin to study the millimeter-wave radar imaging technique,nevertheless,it suffers seriously from the low resolution,the low robustness of imaging algorithm,and so on.The dissertation aims holistically to research MIMO radar imaging under different orthogonal waveforms in the automotive radar application,and uses some hot topics,sparse array,sparse scattering in space etc.,to analyze the MIMO radar imaging model.Aiming at the above existing problems in MIMO radar imaging,the dissertation firstly combines the sparsity of scattering and the sparse array to set a more holistic signal model.With the help of the alternating direction model of multipliers(ADMM)and another mathematic method,we have established the MIMO imaging model based on the sparse prior condition.According to the characteristic of the block distribution in space scattering,the dissertation establishes the imaging method,and the reconstruction of compressive sensing(CS)measure matrix is built up.Finally,the amplitude and phase error exist widely in array radar,and further a robustness imaging algorithm is derived based on the characteristic of the error.The concrete research contents can be seen below:1.Focusing on the existing problem of time-division multiplexing MIMO(TDM-MIMO)radar,low resolution and serious Doppler ambiguity,the dissertation has proposed an imaging algorithm based on the reconstruction of sparse arrays.In TDM-MIMO radar,the more the number of transmitters is,and bigger the Doppler ambiguity is.Besides,there is main lobe shifting from the real target position and sidelobe high,caused by switch delay in TDM-MIMO radar.In this dissertation,the sparse array is designed to reduce transceivers,and then an imaging model,combining the low-rankness of the snapshot matrix and the sparsity of scattering points,is established.The sparse location of scattering in points and uniform snapshot can be obtained by minimizing l1 and schatten-p norm,and the simulated data is used to verify the effectiveness of the proposed algorithm.2.The dissertation has proposed a MIMO radar imaging algorithm based on the hybrid sparse array,both the transmitting array and receiving array being a sparse array,aiming to the demand of Doppler-division multiplexing MIMO(DDM-MIMO)radar.There is a serious Doppler ambiguity when the number of transmitters is large,which makes ones design a sparse array to alleviate Doppler ambiguity.Based on the limitation of 4 chirp cascade in the industry application,a hybrid sparse array is designed to expand the array aperture under the same number of transceivers.At last,an imaging algorithm is derived,with a resort to matrix factorization(MF)alleviating computational complexity,and the sparse location and uniform snapshot matrix can be obtained.Compared to the previous algorithm,the algorithm not only can reduce the computational complexity of the algorithms,but also can improve the sparse reconstruction precision.3.The dissertation has proposed a radar super-resolution algorithm according to the characteristic of block distribution in scattering space,under the application background of automotive code-division multiplexing MIMO(CDM-MIMO)radar.The measurement matrix is divided into several subchannels,and the contributing them can be selected by a greedy algorithm.A new measurement matrix can be obtained by setting the non-contributing subchannels zeros,and finally a sparse recovery algorithm can be got by the new measurement matrix.Through the simulation experiments,the proposed algorithm can well inhibit the impact of background noise on imaging results.4.The amplitude and phase error are considered in the MIMO radar imaging.An optimization model of minimizing the amplitude and phase error has been proposed by the sensitivity of lp norm on the uniform noise.However,the mathematical model is non-convex.Facing the problem,a new threshold iteration algorithm is proposed,and by using the algorithm,an exact recovery algorithm and an inexact recovery algorithm are proposed,meanwhile the computational complex and the converge curve of two algorithms are obtained.It concludes that the exact recovery algorithm can achieve a higher recovery precision compared to the inexact recovery algorithm.