Range Doppler detection for automotive FMCW radars

The FMCW-radar-principle is widely used for automotive radar systems. The basic idea for FMCW-radars is to generate a linear frequency ramp as transmit signal. The difference frequency between the transmitted and received signal is determined after downconversion. In order to detect range and velocity together, the information, extracted from one frequency ramp, is not enough, because it is ambiguous. Several subsequent ramps have to be generated to remove the ambiguity between the frequency portion produced by range and the Doppler frequency. In this paper two methods are presented to perform this basic signal processing step. The two methods have the Fast Fourier Transform as basic calculation step in common, but the number of FFTs and their length are different. The requirements on bandwidth for the IF-Hardware and A/D-Converters are also determined by the algorithm for Range-Doppler-Detection. The CFAR (constant false alarm rate)-algorithm for target detection must also be adapted to the chosen method. Both methods have been verified with a 24 GHz radar prototype. The radar-frontend has been built with a newly developed SiGe-radar-chipset on inflneon's B7HF200-process with a transition frequency of 200 GHz.