| In recent years, bistatic multiple-input multiple-output (MIMO) radar has beendrawn more and more attention from researchers and engineers. As a combination ofbistatic radar and MIMO radar, bistatic MIMO radar has the potential advantages bothof bistatic radar, such as reduced space loss, covert operation, and reduced susceptibilityto jamming, and of MIMO radar, such as additional spatial degrees of freedom (DoFs).Also, bistatic MIMO radar has the particular advantage of being able to obtain thetransmit angle information by processing the received data. By exploting theseproperties, the performance of parameter estimation and clutter suppression for airbornebistatic MIMO radar can be improved greatly. This thesis considers some issues inbistatic MIMO radars, including target parameter estimation, waveform design, cluttercharacteristic, and space-time adaptive processing (STAP). The main contributions ofthis thesis are summarized as follows:1．The autocorrelation and corsscorrelation properties of transmit waveform setshave great effects on the performance of MIMO radar. However, it is difficult to designwaveform sets which have ideal autocorrelation property as well as crosscorrelation one.In order to solve this problem, waveform de-correlation method is proposed to cancelthe effects of auto-correlation and cross-correlation of transmit waveform of bistaticMIMO radar in receiver. A direct matrix inverse (DMI) based de-correlation filter isfirst presented by formulating the waveform correlation model in bistatic MIMO radars.Then, the de-correlation filters are derived by using the minimum mean-square error(MMSE) method to reduce the influences of colored noise enhancement. Furthermore, aTaylor polynomial expansion based waveform de-correlation method is used to simplifythe computation complexity of the proposed DMI and MMSE de-correlation filters.Finally, the performance of waveform correlation cancellation is verified by applying itto joint DOD and DOA estimation for bistatic MIMO radar multi-target localization.Simulation results demonstrate the effectiveness of the proposed algorithm.2. In conventional airborne bistatic phased-array radar, range-dependencecompensation is necessary for a two dimensional (2D) STAP filter. By using the MIMOtechniques, airborne bistatic radar can enable the use of both transmit and receive angleinformation to achieve the three-dimensional (3D) STAP. The characteristic of3Dclutter ridge of bistatic MIMO radar is first investigated. It is shown that all the clutterridges are located in the same plane in the3D space when both transmitter and receiverare side-looking. Then, as long as the sum of the target radial velocities relative to the transmitter and the receiver is unequal to zero, the target will not be located in this plane.Thereby, bistatic MIMO radar range-dependent clutter suppression can be achieved by a3D-STAP filter that places a notch the entire clutter plane. Based on this, a3D-mDTand a3D localized reduced-dimension (3D-JDL) method are proposed to improve thesmall support performance of the bistatic MIMO radar, respectively. Simulation resultsdemonstrate the effectiveness of the proposed algorithms.3. The issue of airborne bistatic side-looking MIMO-STAP radar using Dopplerdivision multiple access (DDMA) waveform for ground clutter mitigation is considered.The signal model of bistatic MIMO-STAP radar that using DDMA waveform is firstderived. DDMA waveform sets are similar as some kinds of post-Doppler processing,where the clutter of each transmit DOF is modulated with different normalized Dopplerfrequency offset. In order to recover the transmitting DOFs, the signal model of bistaticMIMO post-Doppler processing is also derived. Since all the clutter ridges are locatedin the same plane in3D space, the modulated clutter plane corresponding to eachtransmit DOF can be sufficiently separated and thus non-overlapping by choosingsuitable normalized Doppler frequency offset and high enough pulse repeat frequency(PRF). Moreover, the slow-time post-Doppler processing can recover the transmitDOFs and realize the orthogonal waveform design. Simulation results show thatrange-dependent clutter can be suppressed effectively by using the DDMA waveformbased bistatic side-looking MIMO radar3D-STAP.4. A3D-projection based range-dependent clutter suppression method is proposed tosolve the lack of homogeneous secondary data in airborne bistatic side-looking MIMOradar. By exploiting the co-planar property of clutter ridges of bistatic side-lookingMIMO radar, the3D clutter ridge can be transformed to a straight line in2D space withproper projection direction, and the range-dependence can be cancelled. Based on this, a3D projection matrix is first designed that maps the received bistatic MIMO clutterridges onto the virtual linear clutter which would be received by monostaticside-looking phased-array. Considering the great computational complexity involved inthe3D projection matrix, a virtual transmit beamforming method is proposed toapproach the3D projection matrix. Furthermore, in order to improve the cluttersuppression performance, two kinds of virtual transmit beamforming design methodsare presented. In the first method, the virtual transmit beamforming weight is optimizedfor maximizing the output signal-interference-noise ratio (SINR) of MIMO-STAP sothat the detection performance can be maximized. The second method consists in finding the closed-form of the optimal virtual transmit beamforming weight by usingthe space-time interpolation based method. Simulation results show that the proposedvirtual transmit beamforming method can approach the3D projection matrix effectivelyto realize the range-dependent clutter suppression. Furthermore, by formulating areduced dimensional STAP filter, fewer range samples are needed for a statisticallystable clutter covariance matrix inversion, thus improving the small sample supportperformance. |
PDF Full Text Request