Abstract—In this paper,we consider the design of planar arrays that optimize direction-of-arrival (DOA) estimation performance. We assume that the single-source DOA is a random variable with a known prior probability distribution, and the sensors of the array are constrained to lie in a region with