Target Localization In Bistatic MIMO Radar System

Multiple-Input Multiple-Output (MIMO) radar with high-resolution spatial spectrum estimationperformance can achieve a flexible transmit diversity, which requires each antenna transmitsorthogonal signals. Bistatic radar not only can extract the angle information of target relative to thereceiving arrays, but also can extract the angle information of target to the emission arrays. So bistaticMIMO radar has good targeting accuracy and high anti-interference ability, and so on.Targetlocalization in bistatic MIMO radar system needs the direction of departure (DOD) and the directionof arrival (DOA) and other information. Target localization in space needs to estimate themultidimensional angles. A simple uniform linear array can not meet this requirement, so we use thecomplex array antenna. Besides, taking into account the real-time targeting, it is necessary to explorethe low computational angle estimation algorithm.At first, a low complexity propagator method (PM) was studied in bistatic MIMO radar system.PM algorithm estimates the propagator by covariance matrix. Then we can get the angles of thetargets, and the angels can automatic pairing at the same time because of the intrinsic link of thedirection of departure and the direction of arrival. PM algorithm has no eigenvalue decomposition onthe covariance matrix and spectral peak searching, so it has low computational. And the performanceof this algorithm is better than the conventional PM algorithm.Besides, for target localization in space, multidimensional angles have to be estimated. A simpleuniform linear array can not meet this requirement, so multidimensional angles estimation algorithmin bistatic MIMO radar system with L array was proposed, such as: estimating signal parameters viarotational invariance techniques (ESPRIT) and PM algorithm. These two algorithms can extract theazimuth and elevation of DOD and the azimuth and elevation of DOA. And these estimated anglescan automatic paired. ESPRIT algorithm estimated the angle of the signal by the rotation invariantfeature of signal subspace. ESPRIT algorithm has high estimation accuracy, but it has to eigenvaluedecomposition on the covariance matrix. Having no eigenvalue decomposition on the covariancematrix, PM algorithm has low computational, and the estimation accuracy of PM is closed to theESPRIT algorithm at the high signal to noise ratio.Finally, Joint angle and Doppler frequency estimation algorithm in bistatic MIMO radar wasstudied. We propose DOA matrix algorithm, ESPRIT algorithm and PM algorithm. Using rotationfactor which generated by time delay, DOA matrix algorithm estimates Doppler frequency at first, and then estimates the direction of departure and the direction of arrival through the least squaresmethod.In contrast, ESPRIT algorithm has the best performance. The performance of DOA matrixalgorithm is better than PM algorithm in the case of less number of antennas. PM algorithm has lowcomputational.