We present progress toward the formulation of a mathematical model for a cognitive communicationnetwork with applications to satellite systems. Our model employs abstract concepts including communicators, communications channels, and demand for capacity. These model elements may be tailored to represent a wide variety of practical communication scenarios. We present a dynamic automated reasoning methodology which uses the model to find communication resource allocations for specific scenarios that are superior to static scheduling approaches.
This reasoning process resolves resource dependencies, enforces communication policies, and learns from previous communication attempts. We have implemented this reasoning process using Answer Set Prolog and used it to plan communications for a constellation of 8 satellites and 3 ground stations. The example demonstrates performance improvement over a static scheduling approach and shows how solutions can be found with reasonable computational effort.