Reduced-Complexity Convex Optimization Method of MIMO Beampattern Synthesis

The beampattern synthesis method based on convex optimization theory can obtain a high-performance beampattern. However, the computational complexity of convex optimization will increase significantly as the number of sensors increases, which presents a great challenge to the application of convex optimization in the multiple-input multiple-output (MIMO) array. This paper decomposed the weight vector into the transmit and receive weight vectors to reduce the dimension of the weight vector. On this basis, the decomposition linear fractional semidefinite relaxation (DLFSDR) algorithm was proposed. Compared with the linear fractional semidefinite relaxation (LFSDR) beampattern synthesis method, the proposed method can significantly reduce the computational complexity of beampattern synthesis for the massive MIMO arrays. Numerical simulations demonstrated the superiority of the proposed approach in terms of computational time.