Optimal Cognitive Beamforming for Target Tracking in MIMO Radar/Sonar

In this paper, a cognitive beamforming method for target tracking by multiple-input multiple-output (MIMO) radar or sonar is proposed. In this method, at each step, the transmit beampattern is sequentially determined based on history observations. The conditional Bayesian Cramér-Rao bound (BCRB) for one-step prediction of the state-vector in target tracking problem was used as the optimization criterion for beampattern design. The proposed method is applied to the problem of target tracking in a shallow underwater environment in the presence of environmental uncertainties. It is shown that the method is able to automatically focus the transmit beampattern toward the target direction within a few steps at very low signal-to-noise ratios (SNRs). The method exhibits much better performance in terms of localization estimation error compared to other methods, such as orthogonal (omnidirectional) transmission.