Satellite Observation Mission Resource Scheduling Based on Dynamic Coalition Algorithm.

This study was conducted in response to the challenges posed by the heterogeneity of ground station resources and the dynamic nature of tasks in satellite observation missions. To combat these issues, we propose a resource scheduling method based on a dynamic coalition algorithm. The method involves constructing a five-dimensional evaluation system including spatial proximity, energy sufficiency, equipment integrity, load balancing, and continuous observation capability, which is combined with an improved simulated annealing algorithm to achieve global optimization of the coalition structure. Then, an energy allocation strategy based on demand is designed to enhance system sustainability. An experiment comparing the greedy, particle swarm, genetic, and simulated annealing algorithms was conducted. The results showed that the task completion rate of the dynamic coalition algorithm reached 93.8%; the resource utilization rate was 85.7%; the energy consumption standard deviation was 18.7; and the convergence speed was 45 iterations for the proposed method. These results were significantly better than those of other algorithms used for comparison. The innovative aspects of this study include ① a dynamic energy allocation model based on normalized priority; ② a simulated annealing optimization framework with hybrid neighborhood operations; and ③ the deep integration of multi-dimensional evaluation metrics and dynamic coalition construction mechanisms. This research provides theoretical support and technical solutions for task scheduling in wireless sensor networks under complex dynamic scenarios.