Pulse Doppler Radar with Antenna Arrays Estimating Angular Velocity

Radar provides information on target distance, direction, and velocity, but not directly on angular velocity. We study signal processing for estimating angular velocity directly by radar. First, we describe the signal model. In addition, the Cramer-Rao lower bound (CRLB) is derived to show the feasibility of angular velocity estimation. Furthermore, a method for estimating angular velocity is proposed in which the angular velocity is obtained by capturing the variations in the angular spectrum after noise reduction for signals represented in two dimensions: array and slow time. Specifically, we focus the signal with a frequency transform in the slow-time direction, insert zeros outside the focused target, and perform an inverse frequency transformation to obtain the signal with less noise. In the array direction, zeros are inserted on both sides, frequency transform is carried out, and interpolation is performed to capture the time variation to the azimuth angle. The angular velocity is obtained by evaluating the captured angular change using the least-squares method. Evaluation results show that angular velocity estimation performance improves with more antennas and higher signal-to-noise power ratio. When inserting zeroes, the appropriate length of target extraction can mitigate the adverse effects of noise and the Gibbs phenomenon. However, the angular velocity estimation performance of the proposed method deviates from that of the CRLB and requires improvement.