Real-Number and Binary Multi-Objective Particle Swarm Optimizations: Aperiodic Antenna Array Designs

摘要

The goal of this paper is to investigate the relationship between different design criteria for aperiodic antenna arrays (the peak sidelobe level, the beamwidth and the element number) using multi-objective particle swarm optimizations (MOPSO). An important feature of aperiodic antenna arrays is that they have a lower peak sidelobe level (SLL) than conventional periodic antenna arrays, if all the elements are uniformly excited. This is achieved by eliminating the periodicity in the array factor. Starting from an iV-element periodic array with a uniform element spacing d, we may keep N and randomize the element locations to form a non-uniform array, or keep d and randomly turn off some elements to form a thinned array. In either case the peak SLL varies with the array configuration in a very nonlinear manner, and a low SLL design can be achieved by applying statistical density tapering [1] or iterative design methodologies [2].