Synthesis of Waveform Covariance Matrix for MIMO Radar Transmit Beampatterns: LASSO and IRLS Approaches

摘要

Multiple-input multiple-output (MIMO) radar systems have a wide range of applications not only in military fields but also in civilian areas. This wide applicability of MIMO radars is due to their improved spatial resolution and flexibility of designing their transmit beampatterns. Motivated by this and the increasing interest in obtaining the optimum beampattern through designing the covariance matrix of the transmit waveform, this paper presents two new algorithms for designing the beampattern of MIMO radar systems through optimizing the covariance matrix for the transmitting waveform. The first algorithm is based on the iteratively reweighted least squares of the deviation errors between the designed and desired beampatterns. The second one depends on the least absolute shrinkage selection operator (LASSO), which shrinks some deviation errors elements and sets others to zero, hence achieving a good trade-off between the sparsity and errors of the devised beampattern. Numerical simulations are carried out to prove the superiority of our proposed algorithms in case of both symmetric and non-symmetric beampatterns.