A new method for the multichannel sound reproduction in a prespecified region

A preprocessing technique for the creation of a desired sound field in a target region using multiple sources is presented. The complex source strengths/weightings are obtained by an optimization in a transformed space formed by a truncated series of radiation modes, and two adaptive algorithms are developed for the optimal filter design. The truncation acts as a filter that eliminates the low gain radiation modes of the sources. Therefore, it is able to improve the efficiency of the array. Compared with conventional optimization methods, the new algorithms work on smaller dimensions with little dependence to the initial values. Simulations show that smaller square error can be achieved under the same input-power constraint.