Low-Cost Antenna Optimization by Means of Gradient Search and Principal Components

The paper discusses low-cost design closure of antenna structures. Our approach exploits trust-region (TR) gradient search with antenna response gradients updated by a sparse application of finite differentiation, limited to the principal components of the Jacobian matrix. Numerical results demonstrate superiority of the method over benchmark procedures: sixty percent savings are achieved compared to the reference algorithm without compromising the design quality.