Distributed satellite cluster (DSC), which is usually used to construct backbone networks for space information networks, can be viewed as a single satellite carrying more payload with stronger ability. To implement traffic grooming in satellite optical backbone networks, the severe rate mismatch between wavelength and a single connection request must be considered due to the scarce characteristic of on‐board wavelength resources. In this paper, we adopt a typical DSC architecture in which the process of traffic grooming can be successfully partitioned into two independent parts, including intracluster traffic grooming and ring traffic grooming. For intracluster traffic grooming, we propose a multimode grooming method and design an improved ant colony algorithm (ACO) to obtain the optimum solutions. Simulation results show that the improved ACO yields better convergence performance than traditional ACO; in addition, the improved ACO has potentials in decreasing the number of wavelengths to be used. For ring traffic grooming, we develop a grooming circle (GC) construction algorithm based on bidirectional graph node partition. The simulation results indicate that the proposed algorithm outperforms greedy algorithm not only in the construction of full GCs but also in improving wavelength utilization efficiency.

中文翻译：

---

为分布式卫星集群网络提供流量疏导

分布式卫星集群（DSC）通常用于构建空间信息网络的骨干网，可以看作是承载更多有效载荷且能力更强的单个卫星。为了在卫星光骨干网中实现流量疏导，由于车载波长资源的稀缺特性，必须考虑波长和单个连接请求之间的严重速率失配。在本文中，我们采用了一种典型的DSC架构，在该架构中，流量疏导过程可以成功地划分为两个独立的部分，包括集群内流量疏导和环形流量疏导。对于集群内流量疏导，我们提出了一种多模式疏导方法，并设计了一种改进的蚁群算法（ACO）以获取最佳解决方案。仿真结果表明，改进后的ACO具有比传统ACO更好的收敛性能。另外，改进的ACO具有减少使用的波长数的潜力。对于环形交通疏导，我们开发了基于双向图节点分区的疏导圈（GC）构造算法。仿真结果表明，该算法不仅在完整的气相色谱仪的构造上优于贪婪算法，而且在提高波长利用效率方面也优于贪婪算法。