Design of Satellite Constellations in Geocentric Orbit

Abstract

When the concept of satellite constellations started to gain importance (the 1970s), it soon became clear that the design of this type of space system was challenging. Also, the sensitivity and dependence of the satellite constellation design parameters on the mission type increased the complexity of the problem. Assuming circular orbits and uniform distribution of satellites introduced simplifications to constellations design like Star and Delta configurations of Walker constellations. These simplifications do not apply when considering a generalized constellation design, i.e., constellations that include different orbital planes with highly elliptical orbits at different altitudes. Various decision variables need optimization in the preliminary design phase of such a system. Nowadays, satellite constellation services are essential in developing the network infrastructure of modern cities and the establishment of accurate local navigation services and global broadband services. This thesis deals with a tool that utilizes a genetic algorithm (GA) to preliminary design a low-cost navigation satellite system that acts as an independent service provider for UAE and the Gulf region. The Pareto solution generated provides alternative options for later decision-making. The GA is employed due to its effectiveness on nonlinear-multi objective problems. For the given application, the tool minimizes the geometric dilution of precision (GDOP) and the total number of satellites. The GDOP is an essential parameter to measure navigation signal accuracy in a given area. The tool can calculate more than 10 different possible optimum solutions with an average of GDOP less than 2 and a total number of satellites close to 35. The total CPU time employed to compute the Pareto solution is approximately one hour.

Date of Award	Apr 2022
Original language	American English