Research On The Algorithms For Deploying Service Function Chains In Data Center Network

With the development of the Internet of Things and cloud computing,the exponential increase in data traffic is expected to inundate data center networks(DCNs).The integration of two emerging paradigms,network function virtualization(NFV)and software-defined networking(SDN)enables internet service providers(ISPs)to deploy service function chains(SFCs)of user requests flexibly while reducing operational and capital expenditures.In the aspect of SFC deployment in data center networks,existing researches have the following shortcomings.Less effort has been directed towards meeting heterogeneous needs,such as high throughput or low latency of user service requests with heterogeneous bandwidth demands in data center networks(DCNs).In addition,considering that the increasing traffic demands characterized by large bandwidth and high burstiness is moving to inter-datacenter elastic optical networks(inter-DC EONs),less effort focuses on the problem of online SFC provisioning in inter-DC EONs with the aim of maximizing ISP profits.The main contents and innovations of this thesis are as follows:(1)To solve the problem of how to deploy SFCs in DCNs to meet the heterogeneous service requirements of heterogeneous user requests,this paper designs an SFC deployment algorithm that considers splitting.The algorithm first determines the splitting strategy and splits user requests with greater bandwidth requirements into several subuser requests.Then this paper proposes a sub-user request deployment algorithm for the purpose of achieving network balancing and saving bandwidth resource consumption.The sub-user request deployment algorithm selects links and nodes based on greedy thoughts to implement SFC deployment.Simulation results show that the algorithm proposed in this paper improves the request acceptance rate of users with large bandwidth requirements and reduces the queuing delay of user requests with low bandwidth requirements.The deployment cost of SFC and network load balancing are also optimized.(2)To solve the problem of online SFC deployment that maximizes the total profit obtained by ISP in inter-DC EON,this paper first proposes a profit model and an integer linear programming model of the problem,and then designs a two-stage SFC deployment algorithm.In the first stage,this paper obtains the candidate paths for SFC deployment based on the K-shortest path algorithm.In the second stage,this paper designs the path weight formula to balance the user request acceptance rate and deployment cost to select the deployment path to maximize the total profit.In addition,in order to improve the utilization rate of spectrum resources on the optical fiber link,another deployment algorithm considering SFC splitting is proposed.The simulation results show that the basic algorithm has good performance in terms of deployment time,user request blocking rate and total profit,and the improved algorithm is expected to further reduce the blocking rate.