Application of Deep Learning in Conformal Phased Array Antennas

Array antenna is a type of antenna composed of many antenna radiation units arranged according to certain rules or randomly, which is widely used in many fields such as communications, radar, navigation and satellites. In this paper, a conformal array antenna with a hemisphere structure is designed. The array consists of 91 antenna elements, each of which is the same left-handed doubly-fed circularly polarized antenna unit. Taking advantage of the fact that deep learning can fit complex nonlinear functions and the computation time is low, an artificial neural network (ANN) is constructed for the array. The trained ANN model can predict the maximum gain at different beam steering angles of the antenna array.