Distributed Satellite Communication System Design: First-Order Interactions between System and Network Architectures

Humanity now exists in the midst of the fast-moving Information Age, a period of history characterized by fast travel and even faster information transfer. As data becomes seemingly more valuable than physical possessions, the introduction of exciting applications for communications services becomes ever more critical for the success – and in some cases, survival – of businesses and even nations. While the majority of these innovations have occurred over cable and fiber, a number of the most socially significant have occurred due to the introduction of satellites. Terrestrial fiber and cable systems have a number of advantages, but the extent of their reach and the cost of installation – in terms of both capital and time – favor industrialized nations over more remote and underdeveloped communities. Even as satellites offer the only real chance for ultimate communications ubiquity and true global unity, there remains a significant cost-benefit barrier. Few commercial satellite systems have succeeded economically without first falling victim to bankruptcy. The upfront capital required to implement a satellite communications system is staggering, and historically satellite companies have failed to adequately match capacity and service options to the current and actual future demand. The design process itself is an inherent limiting factor to the achievable cost and performance of a system. Traditionally, the first step toward designing satellite communication systems – as well as terrestrial, sensor web, and ad hoc networks – has been to specify the system topology (e.g., the orbits of the satellites and the locations of the ground stations) based on the desired market and then to design the network protocols to make the most of the available resources. Such a sequential process assumes that the design of the network architecture (e.g., protocols, packet structure, etc) does not drive the design of the system architecture (e.g., constellation topology, spacecraft design, etc). This thesis will show that in the case of Ka-band distributed satellite communication systems this fundamental assumption is not valid, and can have a significant impact on the success (cost, capacity, customer satisfaction) of the resulting satellite communication system. Furthermore, this thesis will show that how a designer values performance during the design and decision process can have a substantial impact on the quality of the design path taken through the trade space of possible joint architectures.