Motion Modeling and State Estimation With Bistatic Range and Doppler Measurements

Using a low-cost multistatic radar system, where a number of receivers report only bistatic range and Doppler measurements, is a key technology of the future of modern radar detection. In this article, the problem of locally processing the bistatic range and Doppler measurements of one single receiver is investigated. The state is formulated in range-Doppler space rather than Cartesian coordinates. It is proven that the bistatic range and bistatic range rate measurements from consecutive three time steps suffice for filter initialization. To address the problem that the analytic expression of the initial state estimate may be absent in the presence of process and measurement noises, a filter initialization method is developed based on the Monte Carlo sampling technique and the particle swarm optimization (PSO) algorithm. The large original search space is divided into several smaller subintervals to alleviate the problem that PSO may trap in the local optimum. A set of unscented Kalman filters is used to produce state estimates, in the framework of a static multiple-model estimator. Numerical experiments demonstrate the effectiveness of the proposed model and method.