Linear time-variant system identification using FMCW radar systems

Automotive radar sensors have been used for measurements of time-varying scenarios for several years and the demands on their performance are increasing continuously. For the development of future signal processing algorithms, it is useful to investigate the problem of linear system identification as a generalized radar scenario. This paper provides a method for the approximation of a wireless transmission channel and derives how such a linear time-variant system can be identified using an FMCW radar. This universal approach enables the introduction of system functions, each characterizing the system's behavior in different time and frequency representation. These system functions of a wireless transmission channel are illustrated by radar measurements of a radially approaching cyclist, which deliver a profound interpretation of the wireless channel.