Synthesize Multiple V/H Directional Beams for High Altitude Platform Station Based on Deep-Learning Algorithm

Abstract

Abstract High-altitude platform stations (HAPS) have captured a great deal of attention as a leading candidate for the coverage extension. As usual, the HAPS station coverage is limited to the circle, whose radius is tens of kilometers. However, in this paper, a concentric circular array (CCA) with 60 antenna elements resonating at 2.1 GHz is proposed for synthesizing multiple vertical/horizontal (V/H) directional beams, affording the path of the road to be covered. A modified version of the gravitational search algorithm and particle swarm optimization (MGSA-PSO) algorithm was used to optimize the suggested antenna's dimensions. The application of a deep neural network (DNN) with a back-propagation algorithm and weighted MGSA-PSO algorithm was used for multi-V/H beamforming patterns directed to cover different road paths by estimating the appropriate feeding phases of the CCA elements. The Earth Explorer (EE) user interface tool is used to create real roads that exist in reality on the MWS-CST program. To verify the validity of the technique, several descriptive scenarios are placed to beam-form the pattern toward the desired covered area, including the rugged desert areas with mountain chains in addition to other roads in forests where trees abound.