Analysis On MIMO Radar Detection And Estimation

Multiple-input multiple-output (MIMO) radar is a radar system that employs multiple transmit antennas to emit specific waveforms and multiple receive antennas to jointly process the returned signals. MIMO radar builds a bridge between the research in radar and communications. MIMO radar has received much attention from researchers since it was proposed. Considering different antenna placements, MIMO radar can be classified into two categories, one with co-located antennas, and the other with widely separated antennas.In this work, multibeam amplitude comparison angle measurement is studied for the MIMO radar with co-located antennas. The effective transmit-receive joint beam enjoys the advantage of narrower beamwidth and lower sidelobes compared with the traditional phased array radar. We derive the pattern of the MIMO radar joint beam, present the formulas for determining the half-power beamwidth and the height of the peak sidelobe, and analyze how to direct the joint beams to form the desired multiple beams for amplitude comparison angle measurement.In the study of the MIMO radar with widely separated antennas, the contributions of this work are:(1) Moving target detection in homogenous clutterWhen a target with a sufficiently high speed presents a very small radial velocity towards the traditional phased array radar, it cannot be distinguished from clutter and thus cannot be detected. The problem is solved by using widely spreaded antennas. We derive the generalized likelihood ratio test moving target detector for centralized MIMO radar, distributed MIMO radar, and phased array radar, and compare their performances. The pros and cons of replacing the centralized MIMO with distributed MIMO radar are discussed. An adaptive version of the MIMO radar moving target detector is also provided.(2) Velocity estimation and antenna placementThe Cramer-Rao bound is derived for the study of velocity estimation performance using MIMO radar. An optimal strategy for MIMO radar velocity estimation is presented. Considering that the antenna placement has significant effects on the MIMO radar estimation performance, we demonstrate the conditions that need to be met to provide the optimal antenna placement.(3) Joint estimation using noncoherent MIMO radarWhen the MIMO radar is employed to jointly estimate the target position and velocity, it is shown that the maximum likelihood (ML) estimator is noncoherent. The Cramer-Rao bound is derived, the mean square error of the ML estimate is analyzed, and a threshold phenomenon is observed. Theoretical analysis and numerical results both show that the ML estimate approaches the actual parameter value as the product of the number of transmit and receive antennas increases. Two approaches are used to derive the noncoherent MIMO radar ambiguity function. A few illustrative examples are also provided.(4) Coherent and noncoherent processingOne of the main differences between coherent and noncoherent processing is whether phase synchronization is required for all the radar transmitters and receivers. Joint position and velocity estimation is studied for both coherent and noncoherent processing, and the resulting estimation performances are compared. It is shown that in the cases where the coherent processing outperforms the noncoherent approach, the coherent processing can be replaced by the noncoherent one with little loss of performance, provided that the product of the number of transmit and receive antennas is large enough. This is useful for practical purpose, which suggests that if the required amount of antennas is affordable, then the noncoherent processing is preferred since it is much easier to realize.