Parameter Estimation Algorithms For MIMO Radar

As a new type of radar, multiple-input multiple-output (MIMO) radar consists of multiple transmitting antennas and multiple receiving antennas. Transmit antennas and receive antennas can be flexibly configured according to system requirements and each transmit antenna can freely choose signal waveform to obtain more degrees of freedom. MIMO radar has received much attention from researchers since it was proposed.This dissertation studies orthogonal waveform design and array signal processing for MIMO radar. The main contributions of this dissertation are as follows:(1) Performance analysis and couple calibration of bistatic MIMO radar systemWith the orthogonal waveforms are transmited by the transmitter of MIMO radar, the problem of parameter identifiability is investigated in the first part. Based on the bistatic MIMO radar system model, an analysis on the parameter identifiability of MIMO radar was given. By exploiting the uncorrelation of reflection coefficient of the targets, the maximum numbers of targets that could be uniquely identified by the MIMO radar is the product of the number of receive and transmit elements minus one. In the following part, the effect of the hybrid bistatic radar configuration on the direction finding performance is considered. A hybrid bistatic radar configuration which constitutes the conventional phased array radar and MIMO radar configuration was proposed. The new configuration could take advantage of both the coherent processing gain provided by the conventional phased array radar and the spatial gain received from MIMO radar. The average Cramer-Rao bound(ACRB) of the estimation was evaluated to assess the estimation performance for both angle of departure and angle of arrival. By measuring the average CRB, the direction finding performances of different configurations were investigated. At last, according to the structural speciality of MIMO radar and the mutual coupling matrix of uniform linear array is a banded symmetric Toeplitz matrix, a calibration algorithm for both transmitting array and receiving array of a bistatic MIMO radar system was proposed. The calibration can be performed without any calibration source and any mutual information. The coupling coefficients vector is computed by finding the solution of a linear constrained quadratic problem, so eigenvalue decomposition is not requisite.(2) Waveform designPractical radar requirements such as unit peak-to-average power ratio and range compression dictate that we use MIMO radar waveforms that have constant modulus and good auto-and cross-correlation properties. We present in this part new computationally efficient zero correlation zone (ZCZ) implementation method. According to the binary ZCZ code in communication and a spread-spectrum radar polyphase code, a new polyphase code with larger ZCZ is synthesized. The main feature of the new code is the absence of sidelobe in ZCZ in the compressed pulse. The configuration of the radar transmitter and receiver for application is presented.(3) Parameters estimationWe present a multiple targets localization and parameter estimation algorithm for a bistatic MIMO radar system. MUSIC estimator is directly employed to estimate the DOAs, the amplitude and phase estimator (APES) is used to derive closed-form solution of the DODs, then RCS can be obtained from solutions of DODs and DOAs via least squares method. The DODs and RCS of targets can be solved in close form, and all the parameters are paired automatically.It is shown how bistatic MIMO radar with uniform linear may be used with the data extension technique and2-D unitary ESPRIT algorithm to estimate the joint DOA and DOD. A closed-form solution of DOA and DOD of targets is obtained. Then, the RCS of targets are estimated by exploiting the solution of DOA and DOD. All the parameters are paired automatically. The algorithm has low computation complexity.The last part of this dissertation shows how estimation of signal parameters via combining the data matrix reconstruction and least squares (LS) ESPRIT be used to estimate both the DOA and DOD in a bistatic MIMO radar system. The proposed algorithm can be effective for any type of signals such as coherent or noncoherent signals. The proposed algorithm can provide improvement precise parameters estimation over ESPRIT for noncoherent signals, especially in the presence of nonuniform noise.