This thesis presents the design, simulation and analysis of a new network based on the Automatic Packet Reporting System (APRS) concept to enhance communications of CubeSat. Created in the 1980s, APRS is arguably one of the world’s largest Radio Frequency (RF) networks maintained by a community of licensed amateur radio volunteers for a variety of terrestrial purposes. Using these established nodes we can create a virtualized ground station network for satellite downlinks. Based on literature reviews of CubeSat communications systems and their orbital elements, as well as direct surveys of the amateur radio community, comprehensive far-field antenna models were created in 4nec2 for both the ground and space sections of the CubeSat downlink. The ground station antenna models include: a directional AMSAT/OSCAR Yagi-Uda, an omnidirectional J-pole, ground plane, and a whip antenna. The CubeSat antenna models used include dipoles, monopoles, and turnstiles. These models were evaluated and then imported into the Analytical Graphics Inc (AGI) Systems Tool Kit (STK) to form a discrete event simulation based on nonhomogeneous WGS84 location data extracted from real world APRS Internet Gateway (I-Gate) packets. All objects within the simulation use realistic antenna models, node locations, propagation models and satellite orbital mechanics. This simulation can act as a suite of generally applicable example satellites models (ExampleSat) for analytical comparisons that include link budgets, access times, differing data rates, antenna designs, orbital altitudes and ground station locations. Furthermore, it represents a North American continental scale RF satellite communications downlink network model. Using the virtual ground station network alters the net probability distribution of the received signal from the spacecraft. Specifically, the thesis compares non-stationary time series analysis methods upon the net received signal. The virtualized ground station network enables the aggregated received signal to appear stationary. This in turn may allow for alternative signals analysis techniques to improve a CubeSat’s downlink performance. The simulation allows us to generate representative received signals from ExampleSat to the ground station nodes for exploratory analysis, utilizing less arbitrary assumptions. This simulation and the set of models enable a more rapid start to trade-off studies for link budget design. This set of ExampleSat downlink models can be used by new designers of small-satellite communications systems. / February 2017