Results of the MIT Space Communication and Navigation architecture study

NASA is currently conducting an architecture study for the next-generation Space Communication and Navigation system. This is an extremely complex problem with a variety of options in terms of band selection (RF, from S-band to Ka-band and beyond, or optical), network type (bent-pipe, circuit-switched, or packet-switched), fractionation strategies (monolithic, mother-daughters, homogeneous fractionation), orbit and constellation design (GEO/MEO/LEO, number of planes, number of satellites per plane), and so forth. When all the combinations are considered, the size of the tradespace grows to several millions of architectures. The ability of these architectures to meet the requirements from different user communities and other stakeholders (e.g., regulators, international partners) needs to be assessed. In this context, a computational tool was developed to enable the exploration of such large space of architectures in terms of both performance and cost. A preliminary version of this tool was presented in a paper last year. This paper describes an updated version of the tool featuring a higher-fidelity, rule-based scheduling algorithm, as well as several modifications in the architecture enumeration and cost models. It also discusses the validation results for the tool using real TDRSS data, as well as the results and sensitivity analyses for several forward-looking scenarios. Particular emphasis is put on families of architectures that are of interest to NASA, namely TDRSS-like architectures, architectures based on hosted payloads, and highly distributed architectures.