Satellite constellations provide the infrastructure to implement some of the most important global services of our times both in civilian and military applications, ranging from telecommunications to global positioning, and to observation systems. Flower Constellations constitute a set of satellite constellations characterized by periodic dynamics. They have been introduced while trying to augment the existing design methodologies for satellite constellations. The dynamics of a Flower Constellation identify a set of implicit rotating reference frames on which the satellites follow the same closed-loop relative trajectory. In particular, when one of these rotating reference frames is “Planet Centered, Planet Fixed”, then all the orbits become compatible (or resonant) with the planet; consequently, the projection of the relative path on the planet results in a repeating ground track. The satellite constellations design methodology currently most utilized is the Walker Delta Pattern or, more generally, Walker Constellations. The set of orbital planes and initial spacecraft positions are represented by a set of only three integers and two real parameters rather than by all the orbital elements; Flower Constellations provide a more general framework in which most of the former restrictions are removed, by allowing the use of resonant elliptical orbits. Flower Constellations can represent hundreds of spacecraft with a set of 6 integers and 5 real parameters only and existing constellations can be easily reproduced. How to design a Flower Constellation to satisfy specific mission requirements is an important problem for promoting the acceptance of this novel concept by the space community. Therefore one of the main goals of this work is that of proposing design techniques that can be applied to satisfy practical mission requirements. The results obtained by applying Global optimization techniques, such as Genetic Algorithms, to some relevant navigation and Earth observation space-based systems show that the Flower Constellations not only are as effective asWalker Constellations, but can also be applied to non-traditional constellation problem domains, such as regional coverage and reconnaissance.

Satellite constellations are an increasingly attractive option for many commercial and military applications. They provide a robust and distributed method of accomplishing the goals of expensive monolithic satellites. Among the many challenges that satellite constellations engender (challenges in control, coordination, disposal, and other areas), refueling is of particular interest because of the many methods one can use to refuel a constellation and the lifetime implications on the satellites. The present work presents a methodology for carrying out peer-to-peer refueling maneuvers within a constellation. Peer-to-peer (P2P) refueling can be of great value both in cases where a satellite unexpectedly consumes more fuel than it was alloted, and as part of a mixed refueling strategy that will include an outside tanker bringing fuel to the constellation. Without considering mixed-refueling, we formulate the peer-to-peer refueling problem as an assignment problem that seeks to guarantee that all satellites will have the fuel they need to be functional until the next refueling, while concurrently minimizing the cost in fuel that the refueling maneuvers entail. The assignment problem is then solved via auctions, which, by virtue of their distributed nature, can easily and effectively be implemented on a constellation without jeopardizing any robustness properties. Taking as a given that the P2P assignment problem has been solved, and that it has produced some matching among fuel deficient and fuel sufficient satellites, we then seek to sequence those prescribed maneuvers in the most effective manner. The idea is that while a constellation can be expected to have some redundancy, enough satellites leaving their assigned orbital slots will eventually make it impossible for the constellation to function. To tackle this problem, we define a wide class of operability conditions, and present three algorithms that intelligently schedule the maneuvers. We then briefly show how combining the matching and scheduling problems yields a complete methodology for organizing P2P satellite refueling operations.

Lunar constellations providing the capabilities of situational awareness and surveillance for future mission operators are analyzed in this study. The use of specialty orbits, such as sun-synchronous and frozen orbits, are analyzed to determine the applicability of these unique orbits. Additionally, altitude and inclination trades are performed to determine the degree to which mission objectives are achieved through ranges of these orbital parameters. Using the analyzed orbits, constellations of varying patterns are developed and surface coverage figures of merit are used to evaluate them. The research concludes with calculation of the yearly cross-track and in-track stationkeeping costs of a representative constellation. This stationkeeping is necessary for preservation of the designed coverage statistics.

Satellite wind vector data is integral to atmospheric models and forecasts. Currently, the limited refresh rate of global wind vector measurement systems makes it difficult to observe diurnal variation of mesoscale processes. Using advancements in the underlying subsystem technologies, new satellite wind scatterometers may be possible that increase temporal resolution, among other performance metrics. I propose a method for parametrically modeling the extreme performance range of a complex system. I use this method to develop a model of the space of possible satellite wind scatterometer designs. I validate the model using point designs of heritage scatterometers. Finally, I present two example concepts for constellations of cooperative satellite wind scatterometers capable of measuring global ocean surface vector winds every hour for the same total cost as a single heritage scatterometer.