Non-uniform Luneburg lens antennas: a design approach based on genetic algorithms

This paper focuses on the optimum synthesis of Luneburg lens antennas using a novel application of the genetic algorithm (GA) optimizer. Since it is desired to simultaneously control both the achievable gain as well as to reduce grating lobe levels, an adaptive cost function is used in the process of the GA optimization. To determine the scattered field of the lens antenna an analytic technique is used based on the dyadic Green's function of the multi-layered spherical shell. It is shown that the radiation performance of the uniform Luneburg lens is significantly improved by designing a non-uniform lens with a low number of shells. The antenna gain is increased and the grating lobes are suppressed considerably. Many useful engineering design guidelines have been established for the optimum design of the lens.