Signal Processing Methods for Millimetrewave FMCW Radar with High Distance and Doppler Resolution

Abstract

Novel FMCW radar concepts for precise distance and velocity measurement are presented. The systems incorporate a SAW reference delay line used to measure the instantaneous phase of the oscillator signal. The algorithms evaluate the specific effects of Doppler frequency shifts in the presence of phase errors such as oscillator phase noise or a nonlinear FM modulation. A method that combines a sweep linearisation with a predistorted modulation signal and a constant phase initerval sampling technique are presented. ARMA-modeling for high resolution spectral estimation is applied in order to obtain high Doppler and distance resolution at a moderate sweep bandwidth. Furthermore, a nonlinear FM approach is shown that particularly uses phase error effects to determine the object speed. Using this technique the optimal FMCW radar distance resolution at a given sweep bandwidth is obtained independently of the Doppler resolution. The functionality of the concepts is demonstrated at millimetrewave frequencies. The paper illustrates, that the combination of advanced signal processing schemes with a cost-effective front-end can provide high performance microwave sensors.