A New Approach to Signal-to-Noise Ratio Estimation in Adaptive Doppler–Kalman Filter for Radar Systems.

Doppler frequency carries the information about the relative velocity of the object regarding the radar antenna. However, since the target maneuvers can be temporary with high intensity, Doppler frequency has to be estimated by using a window function in the frame of time-dependent Fourier transform. In order to minimize the estimation error, the window function width should be adaptive. The window length adaptation has been performed based on the estimates of target acceleration and signal-to-noise ratio (SNR). In this paper, we give focus on the SNR estimation and propose a new approach based on autoregressive method of spectral estimation, where in one step the amplitude of the sinusoidal signal and noise variance are estimated. The simulation results justify the advantage of the proposed approach in general applications of signal processing of one sinusoid in white noise environment. [ABSTRACT FROM AUTHOR]