Beampattern synthesis with nulling for sub-arrayed MIMO radar under constant modulus constraints

Abstract

Sub-arrayed multiple-input multiple-output (MIMO) radar holds the benefits of waveform diversity and coherent processing gain jointly. Based on prior knowledge, we exploit the sub-arrayed MIMO radar to form multiple beams in the desired directions and null the power toward the interference or clutter. Meanwhile, to maximize the radiation efficiency of nonlinear amplifiers in radar systems, the transmit weight vectors are required to have constant modulus properties, which leads to a complicated nonconvex problem. To that end, an iteration method with high calculation efficiency based on the alternating direction penalty method (ADPM) framework is presented to resolve this problem. The original problem is transformed into a general bi-convex problem with bi-affine constraints via introducing auxiliary variables. A high convergence rate can be achieved by selectively increasing the penalty parameters in each iteration. For any initialization, the ADPM-based method is proved to converge. And the obtained convergent solution fulfills the Karush-Kuhn-Tucker (KKT) conditions. Finally, the effectiveness of the proposed method is demonstrated by simulation results.