| In traditional networks,Network functions are realized by special hardware equipment.Network middleware development and maintenance consume a lot of costs,and deployment is not flexible enough.Network Function Vitualization(NFV)is considered an attractive solution.NFV uses virtualization technology to decouple network functions from dedicated hardware devices to form a virtual network function(VNF)that can be deployed on general-purpose hardware devices.Service providers can deploy virtual network functions flexibly and dynamically.Network operators can reduce costs effectively.Under the NFV system,network services are implemented through a set of VNF combinations with order constraints.The chain composed of these VNFs is called the Service Function Chain(SFC).The focus of the Virtual Network Function Scheduling(VNF-S)research is to determine the execution order of VNFs on each node,and to minimize the total execution time of all SFCs with in order constraints.At present,there are few relative researches on VNF-S issues in this area,and most of the researches focus on traditional unicast services.In view of the insufficient research on multicast VNF-S,the research of this thesis is mainly divided into the following two parts:1)Aiming at the static scenario,a multicast VNF-S problem model in the static scenario is proposed.For multicast services that already exist in the network,network system reasonably schedule the running sequence of VNFs in the network to minimize the maximum completion time of all network services in the network.An improved water wave optimization algorithm is proposed to solve the static multicast VNF-S problem.The algorithm adjusts the coding strategy of the algorithm for the virtual network function scheduling problem,making it more suitable for solving the multicast VNF-S problem.Simulation experiments prove that the improved water wave optimization algorithm has better results.2)Aiming at dynamic scenarios,a multicast VNF-S problem model in dynamic scenarios is proposed.For newly arrived service requests,the maximum completion time of all network services in the network with in the constraints placed by the VNF should be reduced as much as possible.An improved swordfish optimization algorithm is proposed to solve the dynamic multicast VNF-S problem.The improved algorithm uses a variety of group strategies to enhance the search capabilities of the algorithm.Through experimental analysis,the improved sailfish optimization algorithm can better solve the dynamic multicast VNF-S problem. |
