MIMO-radar Waveform Covariance Matrices for High SINR and Low Side-lobe Levels

Abstract

MIMO-radar has better parametric identifiability but compared to phased-array radar it shows loss in signal-to-noise ratio due to non-coherent processing. To exploit the benefits of both MIMO-radar and phased-array two transmit covariance matrices are found. Both of the covariance matrices yield gain in signal-to-interference-plus-noise ratio (SINR) compared to MIMO-radar and have lower side-lobe levels (SLL)'s compared to phased-array and MIMO-radar. Moreover, in contrast to recently introduced phased-MIMO scheme, where each antenna transmit different power, our proposed schemes allows same power transmission from each antenna. The SLL's of the proposed first covariance matrix are higher than the phased-MIMO scheme while the SLL's of the second proposed covariance matrix are lower than the phased-MIMO scheme. The first covariance matrix is generated using an auto-regressive process, which allow us to change the SINR and side lobe levels by changing the auto-regressive parameter, while to generate the second covariance matrix the values of sine function between 0 and $\pi$ with the step size of $\pi/n_T$ are used to form a positive-semidefinite Toeplitiz matrix, where $n_T$ is the number of transmit antennas. Simulation results validate our analytical results.