Design of sparse linear arrays by Monte Carlo importance sampling

The formation of acoustic images in real-time requires an enormous computational burden. To alleviate this demand the use of sparse arrays for beamforming is mandated. The design of these arrays for adequate main-lobe width and low sidelobe level is a difficult nonlinear optimization problem. A new approach to the joint optimization of sensor placement and shading weights is discussed. Based on the concept of importance sampling the optimization method is presented and some examples given to illustrate its effectiveness.