Imaging Experimental System Design And Algorithm Implementation For MIMO Radar

Since MIMO techniques have made a huge success in communication field, the idea of MIMO radar was presented at the beginning of the 21st century. Benifit from the characteristic that MIMO radar transmits orthogonal waveforms and receives the reflected signals in similar ways, an equivalent virtual array is synthesized and many advantages (e.g.,higher angular resolution, better identifiability, etc.) can be obtained. MIMO radar imaging technology, which drawn a lot of attention of reseachers from both home and abroad, is the combination of imaging process in traditional wideband radar and equivalent virtual array.Along with the gradual perfection of the MIMO imaging radar theory system, a experimental system or prototype for actual performance verification become an active demand. For this reason, the theory evidence and implementation procedure of a MIMO radar experimental system as well as the algorithm implementation based on the system are detailedly discribed in this paper.Firstly, the basic theory of MIMO radar imaging is introduced. The beamwidth and angular resolution of attenna array under the far field signal DOA mathematic model are analysized. The mathematic model of FMCW MIMO radar is aslo presented, the methods of angle estimation and imaging under this model are discribed.Secondly, the design procedure of a MIMO imaging radar prototype is described in detail.With the purpose to to minimize the error sources from the radar transmitter, a DDS and PLL structure is used as a FMCW signal generator in the radar transceiver to avoid the appearance of image frequencies caused by mixing. In order to expand angular resolution with limited number of antenna elements and data channels, panel directional antennas accompanied with wide element spacing scheme are applied in the MIMO array design. To restrain out band noise and the coupling effect between transmit and receive antennas, an adjustable range gate which is made up of switched capacitor filters is used in the IF signal conditioning circuit. In order to provide radar images in real-time, a DSP and FPGA based signal processing board is implemented and emploied.Thirdly, beamforming imaging algorithm as well as its improved algorithm applied in FMCW MIMO radar are introduced. The imaging algorithm implementation on the signal prodessing board and the actual radar performance in field experiment is also discribed.Aiming at the problem that false alarms caused by the grating lobes of traditional beamforming algorithm appear in the angular dimension of radar's imaging results, an improved imaging algorithm based on azimuth threshold is proposed. In order to improve the execution efficiency of the algorithm as well as the imaging speed, according to the algorithm request, a highly optimized pipeline based on DSP and FPGA is designed to prodece radar image. With the purpose to evaluate the real performance of the radar, imaging experiment takes place on a square and the experiment result is analysized.Finally, the former content of this paper is expanded by describing the use of super resolution imaging algorithm in FMCW MIMO radar. Based on the theory of subspace and space time equivalence, the methods of range estimation and angle estimation in FMCW MIMO radar with MUSIC algorithm are deduced. By exploiting matrix smoothing technique,spatial smoothing method and time smoothing method are used to decorrelate the targets with angle information and range information respectivly. By using space-time two dimension smoothing method,a data array reconstruction matrix and a new steering vector which contains both angle information and range information of the targets are generated. And then,the two dimension MUSIC algorithm is applied to process the measured data so as to realize super resolution imaging on both range and angle.