Efficient Newton-based pattern synthesis of reconfigurable parasitic arrays

Dense reconfigurable parasitic arrays have many applications including beamforming, null-steering, adaptive matching and frequency agility. Due to the non-convex nature of the objective function, global search methods, such as genetic algorithms and particle swarm optimization have traditionally been employed. However, the computational cost of such algorithms may be prohibitive for dynamic in-situ optimization. It was previously demonstrated that a very efficient method, consisting of a geometric solution followed by Newton-based optimization, produced moderate (suboptimal) main-beam gain and specified pattern nulls. Herein a nearly optimal, yet efficient Newton-based method is developed for dense parasitic arrays based on subspace separation of beam and null optimization. The suitability of the optimization algorithm is demonstrated via numerical examples.