Robust DFT-based signal processing in Micro-Doppler radars

Micro-Doppler (mD) radar is an efficient tool for surveillance systems and they are rapidly developing now [1]. One reason for this is the potential ability of mD radars to provide additional information (features) useful for recognition of objects extracted from short-time spectral or higher-order estimates [2]. In particular, it is known that a signal reflected from a walking human represents a set of harmonics reflected by different parts of a human body. These harmonics have amplitudes and frequencies changing in time. Joint analysis of these frequencies and their evolution in time serve as aforementioned features that are to be retrieved.