Research On Virtual Network Mapping Algorithm Based On Path Pressure And Energy Consumption Awareness

Network virtualization technology is the key technology to solve the "rigid" problem of the Internet and represents the future development direction of the Future Network.Network virtualization is the best way to maximize resource management and utilization in the cloud computing environment.Its core is the problem of virtual network mapping.How to effectively use the underlying network resources in the data center and improve the efficiency of network communications is a research focus.With the deteriorating global environment and increasing energy consumption,the issue of network energy consumption has become a hot topic in the field of network research.Therefore,it is of great significance to carry out research on green virtual network mapping technology.This thesis makes deep research on the virtual network mapping algorithm from two aspects: mapping cost and mapping new energy consumption.The specific work done is as follows:This thesis analyzes the development history of network virtualization and related technologies,and studies the mathematical model of virtual network mapping and the necessary evaluation indicators.For the virtual network mapping algorithm,classification was performed according to different standards.This thesis designs a virtual network mapping algorithm based on node proximity information and path pressure awareness in the virtual network mapping problem with cost priority as the goal.For the traditional algorithm which does not consider the balance of node and link resource consumption,and the problem of the neighboring node resource is not considered in the virtual network mapping process,a two-stage NPIP algorithm is proposed.First,the BFS search ordering is performed on the virtual nodes with the highest total resources,and then the richness of the node resources and adjacent node resources of the candidate physical node set is considered.Then,the best mapping node is selected according to the node fitness value.At the same time,define the threshold of physical resource usage to balance the network node load.The path average pressure is taken into account for the candidate paths that are solved,and the path with the greatest fitness degree is mapped to the virtual link to balance the link load.The simulation results show that the NPIP algorithm can effectively improve the virtual network request acceptance rate and balance the load of the underlying network.In the virtual network mapping problem with energy saving priority as the goal,the thesis considers energy consumption factors and designs a virtual network mapping algorithm based on energy-aware load balancing.The energy consumption model of the virtual network mapping is given.The energy-aware virtual network mapping mixed-integer linear programming model(MILP)is designed.The greedy heuristic algorithm EALB-VNE is used to solve the model.Calculate the relative importance of the candidate physical node set of the virtual node to be mapped.Then,according to the new power consumption of the virtual node mapping on the physical node,define the energy-saving fitness function of the node and select the physical node of the optimal fitness value..When physical candidate paths are selected,the relationship between bandwidth resource consumption and link energy consumption is balanced to achieve the purpose of balancing physical link pressure.The simulation results show that the algorithm can obviously reduce the power consumption of the underlying network and achieve the physical network energy-saving effect.