Evaluation of a Graph Coloring based Satellite Resource Allocation Scheme

This paper evaluates resource allocation (RA) schemes for low Earth orbit (LEO) satellite systems. In traditional multibeam satellite systems with frequency reuse, severe spectral inefficiency can arise when traffic distributions become highly dynamic and heterogeneous, a situation often encountered in LEO networks. In such scenarios, multibeam systems with frequency overlap can be an effective approach, at the cost of increased computational complexity for determining the optimal bandwidth and power allocation. To address this challenge, this paper proposes an RA method based on graph coloring theory, which offers an efficient means of reducing the computational complexity involved in finding optimal allocations. Simulation results demonstrate that the proposed scheme not only lowers computational complexity but also improves performance compared with conventional approaches.