Array pattern synthesis using PBIL

Abstract

This paper describes the application of the PopulationRased Incremental Learning (PBIL) algorithm to array antenna pattern (power) synthesis. PBIL, an optimization algorithm which has similarities t o genetic algorithms (GAS): offers many advantages over the more established methods of pattern synthesis, especially where constraints are imposed on the element excitations. In this paper, we apply PBIL to the synthesis of a cosec2 antenna pattern using an equi-spaced linear array and show the effect of imposin, in-' creasingly severe constraints on the element excitations and on the pattern itself.

The Multicomputer Technology Initiative at UCT

“…The test algorithm that has been implemented in order to make a first comparison of PVM and MPICH performance is an implementation of Population Based Incremental Learning (PBIL) for antenna array design [8]. PBIL is a stochastic search algorithm and uses an iterative 'generation-based' approach to converge on a suitable solution.…”