Resource Optimization for Multi-Satellite Collaborative Networks With Heterogeneous and Differentiated Services in Highly Dynamic Scenarios

High-dynamic hybrid optical–electrical multi-satellite collaborative networks have been identified as a significant direction for the future development of satellite communication systems. However, the complexity and heterogeneity of service scenarios in such networks impose substantial challenges on resource allocation and Quality of Service (QoS) assurance. To address these issues, a multi-objective constrained resource allocation and traffic optimization algorithm is proposed, aiming to achieve coordinated optimization of bandwidth, latency, and reliability. A differentiated QoS model is first constructed based on service types. This model is then integrated with dynamic network topology and real-time link conditions, including link load, available bandwidth, and propagation delay, to formulate a multi-objective optimization problem. Subsequently, an adaptive routing strategy and resource scheduling mechanism are designed. These mechanisms are intended to meet various QoS requirements while enhancing network throughput, reducing delay, and improving resource utilization. Simulation results demonstrate that the proposed algorithm effectively balances multiple objectives under dynamic network conditions. Compared with traditional approaches, significant improvements are observed in terms of service completion rate, latency-sensitive traffic assurance, and load balancing. The proposed solution provides a feasible approach to intelligent resource management in hybrid optical–electrical high-dynamic satellite networks.