An Online Convex Optimization-based Joint Beam Hopping and Resource Allocation Scheme for LEO Satellite Networks

In Low Earth Orbit (LEO) satellite communication networks, effective resource allocation to improve satellite resource utilization is essential. A flexible and practical solution has been found in the use of beam hopping technology. However, there are uncertainties in the parameters of the dynamic LEO satellite network, thus proposing a comprehensive online solution is crucial for coordinating the beam hopping resources. This study employs online convex optimization for resource allocation, addressing time-varying objectives and constraints. To manage the system’s complexity from multivariate coupling, we decompose the original problem into two subproblems. The first optimizes user associations, subcarrier assignments, and power allocations to maximize system throughput, considering uncertain link bandwidths and varying quality of service requirements. We use an online saddle point algorithm with gradient-free stochastic estimation to solve the online resource allocation challenge. Performance is evaluated using dynamic regret and cumulative constraint violation metrics, with upper bounds $\mathcal{O}(\sqrt T {{\Delta }}(T))$ and $\mathcal{O}\left( {{T^{\frac{3}{4}}}{{\Delta }}{{(T)}^{\frac{1}{4}}}} \right)$, proving sublinear results if the drift remains sublinear. The second uses the Simulated Annealing algorithm to solve the illuminated beam selection problem. The effectiveness of the solution is verified through simulation experiments.