Optimization of Virtual Mobility Management based on NFV and SFC

The networking architecture of 5G and various service requirements brings a huge challenge to the existing centralized mobility management due to the deployment of core functional modules in the core network. vMME is proposed as an alternative solution to meet different application demands. Among vMME mapping solutions, 1:N mapping is a prospective way for distributed mobility management. In this paper, the optimization of virtualized mobility management is investigated by using network function virtualization and service function chain for 1:3 mapping of vMME. A general signaling processing flow of vMME based on service function chain is analyzed. The optimization of vMME is formulated as an optimization problem to place chained VNFs of vMME and optimize the performance of vMME, by choosing the total signaling communication overhead cost and migration cost of state data as optimization objectives. Two heuristic algorithms including TSCMA and DMCMA are proposed and evaluated by simulation, and the performance of different VNF composition solutions under various application scenario parameters are presented.