Design of Constellations for GNSS Reflectometry Mission Using the Multiobjective Evolutionary Algorithms

Abstract

In this simulation study, the operational GNSS satellites of Global Positioning System (GPS), Galileo, GLONASS, and BDS, which are currently in service, are used to transmit signals for GNSS Reflectometry (GNSS-R) measurement. LEO constellations composed of 8, 16 and 24 satellites and with two different patterns, the 2D-lattice flower constellation (2D-LFC) and the 3D-lattice flower constellation (3D-LFC), are designed considering the trade-off among three objectives, namely the visited coverage (<i>V C</i>), the revisited coverage (<i>RC</i>) and the total cost of the constellation. Two multi-objective evolution-ary algorithms (MOEAs), the non-dominated sorting genetic algorithm II (NSGA-II) and the multi-objective evolutionary algorithm based on decomposition (MOEA/D), are applied to solve this multi-objective optimization problem (MOP). The optimal constellations meeting the best trade-off for the three objectives are picked out, and the distributions of the reflected points observed by them are presented and compared. It is found that NSGA-II generally performs better with respect to the convergence and the diversity of the Pareto solutions. The optimal trade-off constellations are generally with inclinations of around 67&#x00B0; to 77&#x00B0; and orbital altitudes of nearly 1000 km. For certain number of satellites, the latitudinal and longitudinal distributions of the number of the reflected points observed by the optimal 2D-LFC and 3D-LFC are highly similar to each other. Moreover, with the resolution of 0.25&#x00B0;&#x00D7;0.25&#x00B0;, the VCs of the optimal 8-satellite and 16-satellite 3D-LFCs reach 58.30% and 79.59%, respectively, and the optimal 24-satellite 2D-LFC and 3D-LFC can achieve an average revisit time of about 11.0 h and 10.2 h, respectively.