Synthesis of linear antenna arrays for radio base stations

This paper presents the application of optimization methods for the synthesis of a linear array for radio base stations of mobile communications systems. The desired pattern shall exhibit squared-cosecant shape, so as to provide uniform distribution of power inside the base station cell and to reduce co-channel interference. The synthesis is performed by a combination of optimization methods: the genetic algorithm is used for the initial global search, then the sequential quadratic programming is applied for local search. This approach allows faster convergence than using only one kind of optimization method. The technique is demonstrated for a linear array of isotropic elements and, subsequently, for an array of microstrip antennas.