New Methods For Target Detection Using MIMO Radar

MIMO radar refers to a novel kind of radar system with multiple transmit and receive antennas. It can be configured with antennas co-located or widely separated. Because of the spatial diversity gain, MIMO radar with widely separated antennas has significant advantages in target detection. Warhead target detection is a key issue in the ballistic missile defense system, which is important for national security. How to correctly identify the true and false warheads is a crucial and difficult issue for the target detection in the anti-missile system. This thesis studies innovations of warhead target detection in MIMO radar with widely separated antennas. The motion of warheads is analyzed and the state space model is developed. Then new distributed algorithms is proposed for warhead target recognition. Furthermore, the antennas distribution optimization is discussed for MIMO radar with widely separated antennas. The main contributions of this paper are as follows:In this paper, we introduce the movement and characteristics of the ballistic missile during the three stages. Then we focus on the micro motion of the warheads in the midcourse phase. The state transition equation of the warhead is built, in which the state vector contains translation and micro motion parameters. MIMO radar is employed to observe the warheads and the corresponding distributed observation equation is formed. The state space model is composed of the state transition and observation equations.The differences of micro motion between the true and false warheads can be reflected in the parameter matrix of the state space model. Thereby the true and false warhead recognition is turned into a composite hypothesis testing problem with unknown and multi-dimensional parameter matrix. For the generalized likelihood ratio test(GLRT), the unknown parameters need to be estimated first, leading to heavy calculation burden and slow detection speed. Therefore a locally optimum unknown direction(LOUD) detection is utilized to detect the small differences between the two hypotheses.Based on the LOUD detector, we consider distributed detection methods to obtain higher detection speed. The distributed configuration and classical algorithms are presented. Then consensus and hard decision based distributed detection algorithms are proposed, both of which are on the basis of the parallel architecture. Numerical experiments are carried out to demonstrate the effectiveness of the two distributed algorithms and the comparison of the detection performance between them is presented.Since the detection performance is affected by the antennas layout of MIMO radar, in this paper, we build an optimization model with the objective to improve the detection performance by adjusting the antennas layout. The simulation results show how the antennas position influences the detection probability. The genetic algorithm(GA) is adopted to optimize the distributed detection performance by designing the receive antennas position of MIMO radar.