Research On Heuristic Multi-strategy Service Function Chain Mapping Algorithm

With the ever-changing of Internet technology and the diversification of network services,using the dedicated physical hardware to provide network services leads to low in both network resource utilization and services ratio of performance-to-cost.The operators have to significantly increase Capital Expenditure(CAPEX)and Operating Expense(OPEX)in network maintenance and upgrading.Network Function Virtualization(NFV)technology is one of the effective ways to solve these problems.As a new method for designing,deploying and managing network services,NFV decouples the realization of network functions from one or more dedicated physical network devices,and transforms such a realization into the component entitled Virtualized Network Function(VNF)running in software,so as to provide users with required network functions in the form of Service Function Chain(SFC).SFC mapping is an important problem to be solved urgently in NFV Resource Allocation(NFV-RA).This thesis completes studying the SFC mapping problem in NFV scenario,which is sponsored by the project entitled “Novel Intelligent Algorithms and Formalization of Resource Optimization in Hybrid Elastic Optical Networks” from national natural science foundation of China.This thesis firstly summarizes the relevant research background,significance and the research status at home and abroad,and then introduces the relevant basic theories involved in SFC mapping,including the overview of NFV-RA problem,SFC mapping principle and algorithms classification.It also presents the functions,module composition and function extension of simulation platform Net2 Plan.The main research content and innovation of the thesis include the following two parts.In the most existing researches of SFC mapping process,VNF nodes mapping and virtual links mapping were processed separately,which makes it difficult to efficiently utilize network resources.However,joint mapping of VNF nodes and virtual links can effectively improve the utilization of network resources and reduce the costs of network services.Facing the problem of SFC mapping with limited network resources capacity and aiming at optimizing the total overheads of network resources,an integer linear programming model of this problem is established.The mechanism for joint mapping of VNF nodes and virtual links is designed,and a node location and resource awarenesses based SFC mapping (NLRASM)algorithm with limited resources capacity is proposed and studied.The software programming to simulate the performance of the proposed algorithm is completed.The simulation results show that the proposed algorithm can effectively reduce the total consumption of network resources compared with the existing algorithms.The diversification and flexibility of user requirements lead to the variety in both number and type of VNFs required by each SFC request.To reduce the total costs of SFC requests deployment in general scenario,an integer linear programming model of SFC mapping is established.Accordingly,an VNF multi-point coordinated SFC mapping algorithm with server placement(VMCSM-SP)is proposed and studied.The algorithm determines the placement node locations of the servers according to the characteristics of the physical network topology,and implements the coordination selection between the locations of multiple candidate mapping nodes of adjacent VNFs and the routing of the corresponding virtual links,so as to select the optimal mapping scheme for each SFC.The software programming to evaluate the performance of the proposed algorithm is performed.Compared with the comparison algorithm,the simulation results indicate that the proposed algorithm can effectively reduce the costs of network bandwidth resources usage,VNF deployment and server node resources usage.Finally,the works of this thesis are summarized,and the research directions that can be developed further are given from three aspects of services scenario,SFC split,and multistage collaborative algorithm.