Colocated MIMO radars using the Sparse Fourier Transform

We consider a MIMO radar system with M_t transmit and M_r receive antennas, which uses uniform linear arrays (ULAs), transmits orthogonal waveforms and implements matched filtering at the receive antennas. Under certain conditions, the receive antennas snapshot can be viewed as a signal containing spatial frequencies, which are directly related to target direction of arrivals. For a small number of targets relative to the product M_t M_r, the snapshot is sparse in the spatial frequency domain, thus we propose to employ the Sparse Fourier Transform (SFT) to estimate the targets. The SFT enables the MIMO radar to operate at the same resolution as an M_tM_r-antenna phased array, while employing only O (K log₂ (M_tM_r/2)) matched filters, where K is an integer larger than and proportional to the number of targets; a typical MIMO radar would require M_t M_r matched filters to achieve the same resolution. Thus, the SFT enables lower operational cost for the MIMO radar system.