Study On Waveform Optimization And Signal Processing For MIMO Radars

Multiple-Input Multiple-Output (MIMO) radar systems illuminate targets by transmit-ting different signals via multiple antennas, and receive target echoes with multiple receiv-ing antennas. There are two types of MIMO radar systems according to the space distances among antennas. One is colocated MIMO radar system and the other is statistical MIMO radar system. The colocated MIMO radar system can be further categorized into monostatic and bistatic MIMO radars. Meanwhile, statistical MIMO radar systems can be further cate-gorized into transmit diversity only MIMO radar and transmit-receive diversity MIMO radar. The colocated MIMO radar system has the advantages of better parameter estimation, better identifiability and more flexible transmitting beampattern. Statistical MIMO radar improves detection performance by exploiting the statistical target angular diversity. The capability of target localization and parameter estimation can also be improved by applying coherent processing scheme. This dissertation studies orthogonal waveform design and array signal processing for MIMO radar. The main contributions of this dissertation are listed as follows:1. Target echoes always span several range bins when high range resolution MIMO radar systems are employed. A novel approach to design zero correlation zone (ZCZ) like polyphase codes based on genetic algorithm is proposed. The proposed method selects the ZCZ sequence set as initial population and optimizes the crossover and mutation operations according to the feature of the ZCZ sequence set. With correlated signals energy being pushed far away the mainlobe, the designed polyphase code has lower auto-and cross-correlation sidelobes near the mainlobe.2. The correlation of target echoes using transmit diversity MIMO radar is analyzed. The echo correlation of multiple targets is analyzed under Swerlingâ… andâ…¡target scat-terering model assumptions. The results indicated that the echoes are uncorrelated un-der Swerlingâ…¡model but correlated under Swerlingâ… model. A condition to decrease the correlation of target echoes via increasing the transmitters is given. The condition is shown to be related to transmitted signals.3. A two-step approach is proposed for constructing minimum redundancy (MR) MIMO radars. In the first step, the total number of antennas is minimized with given vir-tual aperture. In the second step, the virtual aperture is maximized by optimizing the configuration of transmit and receive antennas and the spacings between antennas. This method reduces the computational load of designing MR MIMO radars. It is shown that the augmented covariance matrix (ACM) of MR MIMO radars has no guarantee of non-negative definition, which may fail Capon estimation methods. An adaptive diagonal loading technique, whose loading level is adjusted adaptively according to the minimum eigenvalue of the ACM, is introduced, and then robust Capon method is achieved.4. Conventional bistatic altitude measurement radars have a problem of complicated sys-tem configuration. A targets altitude measurement method using virtual elements tech-nique by exploiting bistatic MIMO radars is proposed. The method implements altitude measurement via estimating the angles of targets respect to transmit and receive arrays. Different from the altitude measurement methods of conventional bistatic radars, the proposed method requires neither time synchronization nor communication between transmit and receive ends, which simplifies the configuration of system. In addition, the angles can be paired automatically when multiple targets exist, and so target ghosts are avoided. In the case of spatial colored noise, an ESPRIT like based DOA estimation method which can cancel the spatial colored noise is given. Exploiting the orthogonality of the colored noise after different orthogonal signal matched filters. This method elim-inates the effect of spatial colored noise on angles estimation and improves the estimate performance.5. A coherent processing technique for locating targets with transmit-receiver (distributed) diversity MIMO radar is studied. In order to enable coherent processing techniques for distributed MIMO radar, dual-frequency transmitters are exploited to eliminate the stochastic phases with complex fading coefficients which are caused by target angular fluctuations. The Cramer-Rao bound (CRB) of target localization accuracy is proved to be inversely proportional to the frequency difference among the transmitted signals. Due to the radar positions uncertainty, the average asymptotic performance of the pro-posed method is proportional to the variance of radar position.