Orthgonal Waveform Design In Large-scale MIMO Radar

Since large-scale MIMO arrays have many antenna elements,it is not suitable for MIMO processing at the array level.Thus,for MIMO Radar it should be processed at the sub-array level.In order to coverage larger airspace simultaneously,it is necessary to design dozens of or even hundreds of orthogonal waveforms for MIMO Radar.However,the more orthogonal channels there are,the worse the orthogonality is,and the computation of optimization method is too large to optimize the ideal result.Therefore,this paper focuses on the waveform design of large-scale MIMO radar,the main work is as follows:(1)This paper introduces the common waveform of MIMO radar,and divides the MIMO waveform into the following four categories: Time Division Multiple Address,Frequency Division Multiple Address,Doppler Diversity Multiple Address,Code Division Multiple Address.The characteristics and advantages of each waveform are analyzed,and the Fast-Time Frequency Division Multiple Address(FDMA)and Fast-Time Code Division Multiple Address(CDMA)waveforms are selected as the designing objects according to the characteristics of large-scale MIMO,such as large number of antenna elements,and long detection distance,fast moving targets.(2)The orthogonal multiple carrier Linear Frequency Modulation(LFM)signal in Fast-Time Frequency Division Multiple Address waveform is studied.The principle for selecting frequency modulation bandwidth and frequency interval of the signal is discussed,and two methods of reducing sidelobe by adding window and spectral correction are studied.Through simulation,it is known that adding window is suitable for single LFM signal when the bandwidth product is greater than 25.And spectral correction is better than the windowing method in reducing the autocorrelation sidelobe.So,spectral correction is more suitable for the bandwidth synthesis signal.In addition,the Distance-Doppler coupling effect of orthogonal LFM signal is analyzed,and two correcting methods are studied: positive and negative LFM Doppler compensation and Doppler filter group compensation.The results show that the positive and negative LFM Doppler compensation algorithm is simple and suitable for searching single target,and the Doppler Filter group is suitable for searching multiple targets,but the direct operation leads to larger computation. If we can acquire the rough range of target speed,then by using the Doppler Filter group,the method will have better compensation effect,and the computation would be smaller.(3)The orthogonal phase code signal in Fast-Time Division Multiple Address is analyzed in detail.Firstly,considering the high computation complexity of genetic algorithm and the poor performance of M sequence,the algebraic method is used to design long and multiple sequences orthogonal phase code.In this method,the linear iterative equation is solved by the inverse recursive multiplication,and the sequence family is obtained by the linear Shift register.The simulation results show that the computational complexity is not high,the cross-correlation performance is better than that of M sequence,and the autocorrelation performance is equivalent to M sequence.In addition,for the phase-code Doppler sensitivity problem,several common correction methods are simulated,and the sliding windows with interpolation Doppler compensation algorithm is proposed for four-phase-code.By dividing the echo signal into several sliding window,and four times squared and two times interpolation,the results show that the method fully compensates the Doppler effect of the pulse of echo signal,and has a better range effect for single target with higher signal-to-noise.