An effective hybrid approach for the optimal synthesis of monopulse antennas

The problem of synthesizing "optimal" sum and difference patterns subject to arbitrary sidelobe bounds by means of a simple feeding network is dealt with. It is shown how a hybrid optimization procedure properly acting on both clustering into subarrays and excitations of these latter greatly improves performances of known feeding architectures with respect to existing procedures. Results of the proposed technique, which significantly outperform existing methods, renew the interest of these simple feeding architectures, and emphasize the effectiveness of this kind of hybrid procedures exploiting the convexity of the problem with respect to a part of the unknowns.