Recently, large-scale and mega low earth orbit (LEO) constellations have become the primary development trends in satellite networks. Considering the complex topology and limited processing capabilities of LEO satellites, traditional distributed routing algorithms with significant flooding overhead are not suitable for the era of mega constellations. In this context, centralized routing technology based on Software Defined Network (SDN) architecture has been widely applied. However, centralized routing technology exhibits a slower response to inter-satellite link (ISL) changes due to longer informing path, which induce that longer time is required to restore traffic in the event of ISL terminal or satellite /hardware faults. To this end, this paper proposes a lightweight Geographic-based Auxiliary Routing Scheme (GARS), adapted to Walker Delta constellation, as a traffic fast restoration strategy for SDN-based centralized routing scheme. This scheme has an ultralow computation complexity ofO(1), enabling the rapid calculation and establishment of temporary paths to restore service during the SDN controller in ground station rerouting. In addition, an optimized geometric routing algorithm with anti-loop mechanism is proposed to enhance the stability and reliability of geometric routing algorithm. Simulation results show that (1) compared with traditional centralized routing system, the routing scheme proposed in this paper shortens interruption time of traffic by at least 65.5% and reduces mean packet loss ratio by at least 67.2% without increasing link load. (2) Compared to other traditional geometric routing algorithms without advanced anti-loop mechanism, GARS with the optimized anti-loop mechanism also effectively ensures the continuity of service, reducing packet loss rate and interruption time by at most 96%.

In the architecture of Digital Twin(DT) optical networks, model management is a key component to ensure efficient network operations. It involves not only precise digital mapping of physical network entities but also includes model integration, dynamic updates, and real-time inference. Considering the diversity and complexity of model types when building an optical network DT, this paper proposes a novel DT model management framework for Optical Transport Network (OTN). This framework enables uploading, training, evaluating, and deploying models to perform inference tasks. It supports both online and offline training to meet diverse business demands. Furthermore, to adapt to continuously changing data and environments, the system integrates an incremental learning technique, which enables continuous updates and optimization of models. Additionally, the framework supports the deployment and inference of mathematical models. Additionally, the components deployed by this framework are entirely managed through containerization technology. This approach ensures portability and enhances the system's fault tolerance and high availability. This framework provides an efficient, user-friendly, and scalable solution for comprehensive model management.

Optical inter-satellite link (ISL) with tens/hundreds Gbps is applied into LEO constellation to overcome the insufficient bandwidth of microwave ISL. Moreover, compared with GEO/MEO constellation, LEO constellation suffers from frequent topology switching because of the lowest orbital-altitude. Topology switching between adjacent topology snapshots is generally done through ISL handover (remove or establish ISL). However, service interruption caused by ISL handover within topology switching is a vital issue to address. After ISL handover occurs, isolated topology and slow route convergence (derived from long ISL delay and forwarding the updated message among hundreds/thousands of satellites) may cause severe service interruption. But considering the fact that time-varying topology is predictable and all the ISL handover could be pre-scheduled for topology management purpose, service interruption problem is investigated from the aspect of ISL handover and routing strategy in this paper. Owing to long establishing time of optical ISL, grouped optical ISL handover scheme is raised to decrease route convergence time during topology switching. Furthermore, predictive-update routing scheme based on Open Shortest Path First (OSPF-PUR) is proposed to minimize route convergence time for each ISL handover. For predictable ISL handover, in OSPF-PUR scheme, forwarding table is refreshed according to locally stored ISL handover information but without flooding. Simulation results illustrate that grouped optical ISL handover scheme reduces route convergence time within topology switching while topology connectivity is guaranteed. The route convergence time of ISL handover is effectively decreased by OSPF-PUR scheme. Besides, for OSPF-PUR scheme, packet loss rate remains unchanged while topology switching frequency increases.

Pre-configured protection cycles (p-cycles) have drawn many attentions due to the fully pre-connected cyclic structures, ring-like speed, mesh-like spare capacity efficiencies and fast switching characteristics. In this paper, we propose a novel removal of loop back (RLB) approach with standard protocol extension (the flooding based distributed cycle pre-configuration (F-DCPC)) to solve the resource inefficiency issue due to long restored paths for symmetric dynamic networks. Meanwhile, distributed cycle pre-configuration (DCPC) is a representative technique for automatic p-cycle configuration. But for the mesh network topology under asymmetric service distribution, it works especially with more network convergence time and larger amount of extra controlling overhead. So in our RLB approach, F-DCPC scheme using generalized multi-protocol label switching (GMPLS) protocol to solve the problems of configurations and reconfigurations with asymmetric digraph in high-speed optical mesh networks is presented. Additionally, to describe the basic idea of traffic splitting strategy in the unidirectional resource distribution, we present the mathematic model with a heuristic algorithm to accelerate resource configuration. We evaluate the performance of F-DCPC utilizing proposed RLB approach with various classical p-cycles enumeration algorithms and allocation strategies by experimental simulations implemented in OPNET modeler, and the simulation results show the effectiveness of the proposed scheme.