The multi-objective optimization of beam pattern in the nonuniform array antenna using the evolutionary algorithm

This paper is about the optimization of beamforming weights in the nonlinear array antenna system. We consider the multi-objective optimization using the evolutionary algorithm. First, we constrain two variables - the width of the main-lobe, the level of the side-lobe. Second, we constrain an additional variable, the energy of beamforming weights, on the former system. In each case, we compute the costs of beamforming weights in each generation. At the end of generations, we can get the Pareto frontier which means the optimal solution set. To demonstrate the performance of the proposed optimization, we compare the results between the proposed algorithm and the conventional window method.