Shaped reflector antenna design using large-scale BFGS

A shaped reflector antenna with single feed which produces a footprint that matches the Chinese territory is designed in this paper. Scalar diffraction theory is used in field analysis process. By applying scalar diffraction, explicit gradient function can be easily obtained from the target function. In antenna synthesis process, Broyden-Fletcher-Goldfarb-Shanno (BFGS) which is a widely used large scale gradient optimization method is applied. The reflector´s surface is optimized at discrete sampling points through BFGS. After optimization, the gain of the shaped reflector antenna reaches above 28 dBi on all prescribed area. Compared with Genetic Algorithm (GA), a typical stochastic optimization method, which usually needs many random trials before getting the final desirable results, BFGS shows better optimization efficiency and predictable optimization time in antenna optimization process due to its guaranteed convergence.