A STEERABLE LEAST SQUARE APPROACH FOR PATTERN SYNTHESIS

The least square method has been widely applied in many fields. However, when the approach is used for antenna array pattern synthesis, it is not excellent. In this paper, the least square method is used to synthesize antenna array pattern, and its performance is reviewed. Then contraposing to the shortcoming of the least square method, a new steerable least square (SLS) method is put forward. For an antenna array whose manifold matrix has been determined, the projection matrix equation can be derived from array manifold matrix easily. In order to get premium solution of array element excitation, a novel projection matrix equation with adjustable matrices is adopted. The results of simulations show that the pattern synthesized by the traditional least square method fits the targeted pattern badly and is worse in the key performance indicators of peak level of side-lobe and null beam level than the targeted pattern; however, the pattern synthesized by the new SLS method fits the targeted pattern well in zero point and local peak distribution and is better in the key performance indicators of peak level of side-lobe and null beam level than the targeted pattern.