A stochastic-gradient method to design optimal space-time codes for MIMO radar detection

We consider a MIMO radar system with widely-spaced transmit/receive antennas and study the problem of designing the signal waveforms transmitted by each source node so as to maximize the detection probability for a fixed false alarm rate under correlated Gaussian clutter. The proposed design procedure relies on a stochastic-gradient search and is general enough to be applied to any statistical target model. Examples are provided to compare the performance of the proposed codes with other known coding schemes.