Mimo Radar Transmit Beampattern Synthe- Sis via Minimizing Sidelobe Level

In multi-input multi-output (MIMO) radar transmit beampattern synthesis, most current literature formulates the problems in steradian space. However, since the beampattern and its parameters are both measured and defined in radian space, from the view point of physical meaning, it will be better to reformulate the problems in radian space rather than in steradian space. In this paper, we propose methods in the radian space to synthesize beampatterns based on minimizing sidelobe level for the two main designs in MIMO radar, i.e., minimum sidelobe beampattern design (MSBD) and beampattern matching design (BMD). For MSBD, the design criteria considering both peak sidelobe level and integrated sidelobe level is proposed. By this we can have a good tradeoff between the intensity and power distribution in beampattern synthesis. After a two-step converting, the formulation of the criteria is transformed into a convex programming, where a global optimal solution can be obtained. For BMD, instead of minimizing mean square error directly as in conventional methods, we propose a power-approximation-based method by minimizing integrated sidelobe level. Finally, numerical comparisons with classical methods demonstrate that the proposed MSBD maintains for all range of main lobe width and the proposed BMD has smoother main lobes with maximal power focused in.