Research On Signal Processing Techniques For Phase-coded MIMO Radar

Motivated by recent developments and attractive advantages of MIMO (multiple-input multiple-output) technology in modern wireless communication field, a new concept of MIMO radar is explored recently, which has become a hot international research area. Two kinds of MIMO radar are put forward. One is statistical MIMO radar in which the transmit and receive antennas are widely spaced. It can resist the RCS scintillation effect encountered in radar systems by capitalizing on the spatial diversity with the improvement of the detection performance. The other is MIMO radar with collocated antennas which employs standard array to transmit multiple probing signals and receive the backscattered signals reflected from the targets. The signals transmitted and received respectively via collocated antennas are fully coherent. MIMO radar with collocated antennas can improve the angular resolution and increase the upper limit on the number of targets which can be detected and localized by the array. MIMO radar could employ code diversity, frequency diversity or both to derive the orthogonal waveform. MIMO radar using code diversity is studied in this paper. The problems of Doppler sensitivity of phase-coded radar and MIMO radar, target detection and parameter estimation of tatistical MIMO radar, angle estimation of MIMO radar with collocated antennas, and high speed moving target detection using MIMO radar are considered herein.The main works and contributions are summarized as follow:1. The performances analysis of pulse compression for phase-coded radar and MIMO radarTake m-sequence phase modulation radar for example, the relation between main-to-sidelobe ratio and Doppler frequency after the pulse compression is studied, and the deduction of its mathematical expression is presented. Since the pulse compression with the non-target range gate in m-sequence phase modulation CW radar is a "wide-band input, narrow-band output process", the sidelobe's real part and imaginary part after the pulse compression approximately obey normal distribution, respectively. Thus, the peak sidelobe's real part and imaginary part after the pulse compression can be calculated. Then the mathematical expression of the main-to-sidelobe ratio after the pulse compression is also gained. The data calculated by the mathematical expression of the main-to-sidelobe ratio is close to the data attained by the simulation, when the period length of m-sequence P≥31. On this basis, the problem of Doppler sensitivity of phase-coded MIMO radar is also studied. It mainly includes the mismatch of pulse compression and orthogonality damage of phase coded signals caused by Doppler frequency. This could provide some help to our subsequent work.2. Non-coherent angle estimation and coherent signal processing for statistical MIMO radar with transmit diversity onlyThe statistical MIMO radar with transmit diversity only can overcome target's radar cross section (RCS) fluctuations by exploring transmit diversity to improve the direction finding performance. Non-coherent angle estimation for statistical MIMO radar with transmit diversity only is discussed, the signal model of which focuses on the effect of the target spatial properties ignoring range and Doppler effects. The received signals of array are processed directly or after pulse compression and data rearrangement for angle estimation. The latter has better performance than the former. Then, based on the complex signal model for the moving and slow fluctuating targets that focuses on the effect of the target spatial properties and range and Doppler frequency, a scheme for coherent signal processing using statistical MIMO radar with transmit diversity only is studied. The echoes of the same target from different transmit antennas are added after the compensations of range delay, Doppler frequency shift and phase. Thus, the highly precise estimation of target azimuth can be obtained from the added target's signal with the high signal noise ratio, which is better than that of Non-coherent angle estimation method.3. Multi-target localization using statistical MIMO radar with receiving diversity onlyAiming at the disadvantages of low positioning accuracy and target ghosts generation in multistatic radar with broad transmit beam and broad receive beam, statistical MIMO radar with receiving diversity only is discussed in this paper. This MIMO radar is a special structure of STMR (single transmitting station and multiple receiving stations)-multistatic radar with broad transmit beam and broad receive beam, and it avoids the complicated beam scan synchronization. A scheme for multi-target localization using Capon space spectrum estimation algorithm in MIMO radar with receiving diversity only is proposed, and the positioning accuracies of the MIMO radar and STMR multistaic radar are compared. Simulations show that statistical MIMO radar with receiving diversity only solves effectively the problems of low positioning accuracy and target ghosts generation in multistatic radar with broad transmit beam and broad receive beam.4. A joint direction of departures (DODs) and direction of arrivals (DOAs) estimation for bistatic MIMO radar with collocated antennas(1) Angle estimation in the presence of spatial Gaussian white noise:A) In order to avoid two-dimension (2-D) angle search for Capon algorithm, ESPRIT algorithm is used for targets'directions estimation with the utilization of the invariance property of the transmit array and the receive array, which decomposes the 2-D angles estimation problem into two independent one-dimensional (1-D) angles estimation problems. Then the interrelationship between the two 1-D ESPRIT is exploited to obtained automatically paired transmit angles and receive angles estimation without debasing the performance of angles estimation. B) Due to the formed virtual array with more elements than the physical array elements, ESPRIT algorithm requires the estimation and SVD (Singular Value Decomposition) of the high dimension covariance matrix of the received signals which are the main computational burden in traditional subspace method. Thus, a fast joint DODs and DOAs estimation using propagator method without the estimation and SVD of the covariance matrix is studied.(2) Angle estimation in the presence of spatial colored noise: A new method for joint DOD and DOA estimation is studied by exploiting the characteristic that the cross-covariance matrix of noises in the cross-correlation matrix of two received data from two transmit subarrays is 0. The DODs and DOAs of targets are estimated via both ESPRIT and SVD of cross-correlation matrix of the received data from two transmit subarrays. The DOAs and DODs of targets can be solved in closed form and paired automatically. It could be effective for three-or more-transmitters configuration system with the influence of spatial colored noise eliminated, and the ability of the spatial colored noise elimination can be improved with the increase of the number of transmitters.5. High speed moving target detection using monostatic MIMO radar with collocated antennasIt is difficult for the traditional radar to detect effectively the high speed moving target with the long time coherent integration method. In this paper, the monostatic MIMO radar with collocated antennas, in which the parallel coherent processing of the multi-channel echo data in a short time is exploited to replace the long time coherent integrative detection, is used for high speed moving target detection. Thus the detection performance of MIMO radar may not be affected by the range migration, the radial velocity variety and the fast Radar Cross-Section (RCS) fluctuation. Meanwhile, MIMO radar with collocated antennas can improve the angular resolution and increase the maximum number of identified targets. Thus, it has the ability to effectively detect a large number of dense high speed moving targets.