FMCW-radar. Signal processing and parameter estimation

This report concerns the development of a real time algorithm to estimate the path parameters of a threat target, for protection of military hardware. The target is assumed to move in a straight line with parameters pass distance, pass time and target velocity. To estimate the path parameters a number of measured ranges and range speeds are needed, which are obtained from one sensor in this study. The goal is to develop a method having high accuracy. The received FMCW-signal, which is delayed by the transit time, is mixed with the transmitted FMCW-signal and analysed in two dimensions for further information. The mixed (filtered) signal consists of two different frequencies i.e. the power is divided between these, which has negative effect on the estimation of range and radial velocity. To avoid this problem and to increase the signal power, one part of the frequency spectrum is shifted, the discontinuity is fixed and the estimate of range and velocity relative to the radar is refined by the periodogram method. The parameters are solved by the Mean Least Square method in two steps and the result is a feasible value for the further search. To optimise the path parameters, Maximum Likelihood method is studied and the minimum variance has been analysed with the Cramér Rao Lower Bound.