DoA Estimation for Automotive FMCW Radar introduces the theory, implementation, and practical validation of direction-of-arrival estimation methods used in automotive radar signal processing. The course begins with phased-array fundamentals and explains how spatial phase differences across antenna elements encode target direction. From this foundation, the main DoA estimation methods are developed and compared: Spatial FFT, Bartlett beamforming, Capon/MVDR, MUSIC, and ESPRIT.
Each method is examined through numerical examples and simulations, with emphasis on its operating principle, angular resolution, assumptions, strengths, and practical limitations. The theoretical part is followed by validation using real measurements acquired with a Texas Instruments AWR2243 automotive FMCW radar configured as a two-transmitter, four-receiver TDM MIMO system forming an eight-element virtual antenna array.
Students work with provided measurement data and Python processing scripts to examine single-target and two-target scenarios and compare the behavior of conventional and super-resolution DoA methods on real radar data.
