Optimization of subarray amplitude tapers

Large phased array antennas are very expensive to build. A way to cut the costs is to make more efficient feed networks. One efficient feed network for a low sidelobe array consists of amplitude tapering at the subarray level while keeping the individual elements in the subarray uniformly weighted. In this way, subarrays could be mass produced and tested. Unfortunately, grating lobes ruin the low sidelobes in the far field pattern. An improvement to this approach is to find an amplitude taper that is constant for each subarray. Thus, all the elements in a subarray are tapered, but the taper is the same (or a mirror image) for each subarray. This approach yielded excellent results for a linear array. This paper improves the previous approach by using a genetic algorithm to design the amplitude tapers for both the elements in the subarrays and the subarray outputs for linear and planar arrays. Because genetic algorithms can optimize a large number of parameters, both linear and planar subarrayed antennas can be optimized.