Research On Network Slicing Technology In Spatial Information Network

Mobile communication has ushered in the 5G era.In the future,communication services will be more abundant and diverse.More communication requirements will be generated in different business scenarios,such as higher requirements for bandwidth resources and the need for new network architectures.The space information network uses synchronous orbit satellites,low-medium orbit satellites,stratospheric balloons,and drones as carriers to acquire,transmit and process space information in real-time.The satellite network,as an essential part of the spatial information network,has the advantages of high bandwidth and comprehensive coverage,and its powerful transmission capacity is very suitable for integration into the ground network as a transmission network.Therefore,it can be used as a supplement to the terrestrial network,seamlessly connecting with the terrestrial network,greatly expanding network resources,and providing more flexible networking functions.At the same time,the satellite network structure is complex,has limited resource and lack flexibility and management capabilities.You can use network slicing related technologies to improve the satellite network structure to realize the centralized management of the satellite network and hide the complexity of the satellite network through the programmable interface.It also divides the satellite transmission network into multiple logical networks through virtualization technology,and responds to the customized needs of users while making full use of resources.This article first proposes a new satellite network slicing architecture.By analyzing the various components of the satellite network,the satellite resources are pooled to form a more flexible satellite transmission network architecture based on SDN.This architecture is easy to interface with the core network and the access network to develop a complete network slice and provide users with differentiated services.A network slice is a logical network.Each sub-slice network provides different network functions and network characteristics,and is instantiated as a network slice instance.The transmission network mainly provides bandwidth resources and forwarding calculation and storage resources.Therefore,when a service request arrives,slice resource mapping is performed according to the service resource request.According to the characteristics of satellite network movement,this paper applies the “snapshot model”of satellite network,and proposes a resource mapping method based on max-min ant colony system within the snapshot time.Compared with the commonly heuristic algorithm,this method has better performance in terms of the success rate of slice request resource mapping and network link load balancing.In addition,in view of the difference in service requests for different snapshot areas during satellite movement,it is necessary to dynamically allocate the resources of the satellite nodes at different snapshot positions.This paper proposes resource allocation algorithms for satellite nodes based on genetic algorithm and reinforcement learning.The resource utilization rate of the system and the cost of node slice resource changes are considered at the same time during allocation.Simulations show that the algorithm proposed in this paper can effectively carry out reasonable resource allocation to satisfy the changes of service requests in different snapshot areas.The reinforcement learning considers the full cycle resource adjustment scheme of satellite nodes,which has better algorithm performance.