LEO (Low Earth Orbit)星座网络凭借其全球覆盖能力,能够有效扩展和补充地面网络的不足,但其高速运动和拓扑结构快速变化为控制器的部署带来了巨大挑战。为应对LEO星座网络中动态拓扑的挑战,本文提出了一种基于区域划分的SDN多控制器选择与动态迁移方法,采用Voronoi区域划分,并引入动态评分机制,综合考虑卫星与种子点的距离、资源使用情况和卫星在区域内停留时间等因素,动态选拔最优控制器。此外,本文提出的控制器迁移策略,通过监测域内平均时延和控制器的资源消耗,确保控制器及时迁移,提升管理效率与稳定性。仿真实验表明,本文提出的SDN多控制器部署方法在降低传输时延方面优于现有方法,同时有效优化了控制器的资源使用率,更加适应LEO星座网络的频繁拓扑变化。The LEO (Low Earth Orbit) constellation network, with its global coverage capability, can effectively expand and supplement the deficiencies of terrestrial networks. However, its high-speed movement and rapidly changing topology pose a huge challenge for the placement of controllers. To address the challenge of dynamic topology in LEO constellation networks, this paper proposes an SDN multi-controller selection and dynamic migration method based on regional division. It adopts Voronoi regional division and introduces a dynamic scoring mechanism. By comprehensively considering factors such as the distance between satellites and seed points, resource usage, and the residence time of satellites in the region to dynamically select the optimal controller. In addition, the controller migration strategy proposed in this paper ensures timely migration of controllers by monitoring the average delay within the domain and the resource consumption of the controller. Simulation experiments show that the SDN multi-controller deployment method proposed in this paper is superior to existing methods in reducing transmission delay. At the same time, it effe
为提高卫星星座网络受到攻击后的抗毁性及工作能力,提出了一种模拟退火狼群算法。该算法利用主客观权重法结合综合逼近理想排序法(TOPSIS:Technique for Order Preference by Similarity to Ideal Solution)对网络中的节点进行重要度评估,并按照节点重要度排序依次攻击。以网络连通度与网络连通效率为优化目标,卫星星座网络通信限制为约束条件,采用运动算子的思想实现狼群自适应步长的游走、召唤和围攻。使用通过优化得出的加边方案对网络结构进行优化。实验表明,与其他优化算法相比,该算法具有优越性,解决了卫星星座网络在受到攻击后工作能力下降的问题,提高了其受到攻击后的抗毁性。
