Network slicing has become an essential technology for the future network. Obviously, it will play an important role in satellite networks as well. To address that quality of service (QoS) may be severely affected by embedding satellite virtual networks (SVNs), we propose a method for SVN admission control that can effectively guarantee the QoS of network slices by admitting SVNs embedded in the physical satellite networks. Specifically, firstly, we propose a two-stage SVN embedding mechanism that decouples short-term resource allocation from long-term admission control and resource leasing. Then, we consider the case of uncertain system capacity due to the highly dynamic nature of the satellite networks topology, and model the admission control problem as a robust optimization problem. The robust problem is transformed into a convex counterpart by using the Bernstein approximation. Finally, we solve the resource allocation problem by converting it into a convex problem. The simulation results show the effectiveness of the proposed method. 

Network Slice Admission Control and Resource Allocation in LEO Satellite Networks: A Robust Optimization Approach

In the satellite ground integrated network, the business types are diverse and Spatio-temporal, and different business types have different deterministic requirements for time delay, timeliness, and other indicators. Network slicing will be a crucial technology to meet the transmission requirements of diversified application scenarios. This paper introduces a dynamic slicing strategy for satellite–ground integrated networks. Firstly, satellite resource mirror nodes are established to mirror satellite information to shield the influence of satellite dynamics on resources. Secondly, the characteristics of the deterministic business are analyzed, and the average satisfaction of long-term slices of the system is taken as the optimization problem. Lastly, we design a algorithms of resource reserved in slices and adaptive adjustment between slices (R2A2) to solve the problem. Simulation results show that the proposed strategy has better performance.

A dynamic Integrated Network Slicing Strategy for Future Networks

Network slicing has become an essential technology for the future network. Obviously, it will play an important role in satellite networks as well. To address that quality of service (QoS) may be severely affected by embedding satellite virtual networks (SVNs), we propose a method for SVN admission control that can effectively guarantee the QoS of network slices by admitting SVNs embedded in the physical satellite networks. Specifically, firstly, we propose a two-stage SVN embedding mechanism that decouples short-term resource allocation from long-term admission control and resource leasing. Then, we consider the case of uncertain system capacity due to the highly dynamic nature of the satellite networks topology, and model the admission control problem as a robust optimization problem. The robust problem is transformed into a convex counterpart by using the Bernstein approximation. Finally, we solve the resource allocation problem by converting it into a convex problem. The simulation results show the effectiveness of the proposed method.

To solve the problems of LEO (Low Earth Orbit) satellite network, such as insufficient flexibility and low resource utilization, an architecture using network slicing technology is proposed. Then a prototype system based on the Docker platform and DTN (Delay Tolerant Network) is implemented, providing IMS (IP Multimedia Subsystem) network slice, Internet of Things network slice and web network slice. Tested by prototype system, the results show that each network slice can independently provide relevant service without affecting each other, and can be allocated network resources flexibly according to the performance requirements of the services. The architecture can effectively improve the flexibility of the core network of LEO satellite network.

Pang Jinkun

Papers

Network Slicing in LEO Satellite Network