Orchestrating for deployment of network slicing

Network slicing has been proposed as an innovative concept to conduct a dedicated and customized logical network with virtualized and physical resources, which help operators to launch and customize service with more flexibility. We analyse the relationship between the network slicing and physical network and define a mathematical model to construct logical network function graphs that can be mapped to the physical network. Further, we focus on demand oriented deployment of Network slicing applied to various application scenarios and formulate it as the integer quadratic programming for the placement of the network functions and connecting them together considering the limited network resources and QoS provision. We mainly discuss three kinds of network slicing deployment policy that highly influence resource cost of physical network. We evaluate resource utilization of network slicing model using a small substrate network scenario and analyze its deployment rule according to a combination of multiple optimization objectives and constraints implemented in YALMIP with CPLEX tools.