The paper presents an approach for target position recovery from frequency modulated continuous wave (FMCW) radar data. The presented algorithm is, in contrast to many established techniques, able to use snapshots taken from arbitrary positions. The presented technique can be applied to imaging as well as localization applications. The basis of the approach is a compressed sensing based algorithm called normalized iterative hard thresholding (NIHT). To prove its capabilities it is evaluated by simulation over a wide range of signal bandwidths and base frequencies.
