An Optimized Frequency Plan Algorithm for Non-geostationary Constellations Using Integer Linear Programming

The unprecedented growth of massive satellite constellations, as well as the advent of a significant number of steerable beams and digital payloads, poses new challenges in determining how to distribute satellite resources. New resource management strategies that operate in high-dimensional and dynamic environments will be required to meet these new challenges. Under the context of satellite communications, the resource allocation (RA) problem is decomposed into six sub-problems: frequency assignment, power allocation, beam placement, user grouping, gateway routing, and satellite routing. Existing conventional techniques of satellite resource allocation become unfeasible to deal with these new challenges and new algorithms have to be developed. The majority of frequency assignment methods fail to fulfill the requirements of the high-dimensional and dynamic environments without defaulting on bandwidth utilization and power efficiency. The work in this paper proposes a new frequency assignment algorithm based on a mathematical programming language (AMPL) that can completely design a dynamic frequency plan with bearing in mind system constraints like handovers and interference. The proposed algorithm is evaluated with multiple objective functions such as bandwidth maximization and produces optimal solutions. Experimentally, the proposed algorithm can allocate at least 155% more bandwidth compared to previous baseline benchmarks.