GAs using varied and fixed binary chromosome lengths and real chromosomes for low sidelobe spherical-circular array pattern synthesis

Low sidelobe array pattern synthesis is achieved by finding appropriate phase shifter settings and amplitude weights for each element. The array pattern synthesis using a genetic algorithm (GA) has been shown in many references, and the various array types are used for the pattern synthesis. The performance of GAs using various types of chromosomes for low sidelobe spherical-circular array pattern synthesis are compared. The spherical-circular array layout consists of eight circular arrays on the surface of a sphere. The GAs control the phase shifters and the amplitude weights of array elements, and minimize the peak sidelobe level while they maintain the mainbeam direction and gain. Three methods are simulated. The first method is a GA with varied binary chromosome lengths (GA-VBCL), and the second method is a GA with a fixed binary chromosome length (GA-FBCL). In addition, the last method is a GA with real continuous chromosomes (GARCC). The convergence speeds and the adapted peak sidelobe levels of these three methods are compared when a good seed (-30 dB amplitude taper) is included/not included in the population.