Optimization and Scheduling Methodologies to Enable Low Earth Orbit Nano-satellite Communication
MetadataShow full item record
Communications with low earth orbit (LEO) nano-satellites (nanosats) are challenging due to the short contact time intervals with ground nodes and uncertainty in successful delivery of messages due to the varying signal-to-noise ratio (SNR) in the environment. This thesis presents optimization models to enable minimum-delay store-and-forward communications between terrestrial gateways and remote users (e.g., ships) via nanosats. Optimization models are formulated to enable timely delivery of messages between nanosats and remote users. Unit-sized messages destined for remote users must be routed from gateways to nanosats to final remote destinations. The connection between nanosats and remote users may not always be well established. The uncertainty in knowing if a message needs to be sent again or was successfully delivered is modeled using a chance constraint in the optimization model. A network flow program is formulated to optimize the scheduling and routing of messages from a central command and control center (CCC) to gateways to nanosats and ultimately to the remote user. The decisions are chosen to minimize the total message delivery time while considering the nanosat contact time windows with gateways and remote users and the solar charging time windows of the nanosats. Although the scheduling and routing decisions are binary variables, the optimization models are shown to satisfy the integrality property. Therefore the relaxed network model can be solved much faster than a binary integer problem. Results on a realistic problem are presented. Comparisons are made to consider the difference between a basic deterministic model, a model with energy constraints, and a chance-constrained model (with and without energy constraints). Comparisons are also made to simple greedy heuristics to demonstrate the value of optimization.