Online Fault-Tolerant Backup Deployment for Virtual Network Functions in LEO Satellite Networks

Low Earth Orbit (LEO) networks have become an important part of the sixth-generation communication system because of the wide coverage and low latency. By introducing network virtualization technology, LEO satellite general servers can be equipped with diversified service functions to improve service deployment flexibility. However, corresponding reliability assurance strategies are needed to cope with inter-satellite link switching and satellite node overload that may lead to corresponding failures. Therefore, we propose a virtual service function backup algorithm based on online fault tolerance in LEO satellite networks. By building a virtual network function backup model for the LEO satellite network, when a network failure or link interruption occurs, the backup instance can be quickly enabled to ensure service continuity. To further optimize the backup strategy and resource scheduling, a Markov decision model is established, and a double-deep Q-network reinforcement learning algorithm is used to dynamically adjust the deployment of backup instances and resource reservation according to the current network node failure conditions. The proposed algorithm is capable of maximizing the reliability and fault tolerance of the system in the face of different types of failure scenarios. The simulation results show that the proposed algorithm improves the success rate of service function chain requests and service survival time under multiple failure scenarios.