Target position and vehicle trajectory estimation with radar signal and data processing

Abstract

Since the wavelength of electromagnetic waves mainly used by radar is considerably longer than that of light, the wave diffracted more easily, and signals reflected from multiple targets are mixed. In addition, while multipath from the transmitter to the receiver is used for data communication, since radar's goal is to estimate the target information (distance, Doppler, angle) from the mixed signal, multipath is excluded. When the radar signal is propagated, the signal strength is inversely proportional to the surface area of the sphere whose radius is the travel distance, so the signal that has undergone multipath is negligible compared to the signal from the line of sight. This dissertation starts with explaining the principles of radar and proposes techniques to overcome the limitations of conventional radar imaging (constant linear motion and side-looking geometry). Subsequently, a method for accurately estimating a target position (especially an angle) using multiple-input multiple-output (MIMO) radar and a target positioning method considering the atmospheric refractivity profile is proposed. Finally, we propose an efficient radar data processing technique for estimating the trajectory of a moving vehicle. For the evaluation of the presented techniques, comparisons with the mean square error (MSE) of the existing techniques as well as with the lower bound of the estimation variance, Cramér-Rao bound or posterior Cramér-Rao bound, are presented.