Research On Key Technologies Of Resource Adaptation For Service Chain In Smart Identifier Network

As one of the most important infrastructures of human society,the Internet plays an irreplaceable role in politics,economy,education and culture.However,the existing Internet architecture has hardly changed for decades.The"static"and"rigid"network architecture has gradually exposed a series of shortcomings such as low resource utilization and poor Quality of Service(QoS).To this end,countries around the world have launched researches on new network architectures.As one innovative future network architecture independently developed by China,Smart Identity NETwork(SINET)aims to solve the"triple binding"problem of the traditional Internet from the root cause and seize the first opportunity as the future of information emerges on the international stage.Based on the main idea of SINET,this paper conducts in-depth researches on the key technologies of service chain resource adaptation,in order to solve the three major issues about how to optimize the latency for services,how to improve the reliability of services and how to improve the energy efficiency of the network.The main contributions and innovations are as introducted follows:(1)To guarantee the end-to-end delay of service chain in SINET,we propose a low latency service chain resource adaptation mechanism.We firstly expound the deployment process of cross-domain service function chain,and the main idea of hierarchical orchestration.Then,considering that the cross-domain service function chain will generate additional network delay,we propose a mathematical optimization method based on integer linear programming,which simplifies the placement problem of cross-domain service chain into two sub-problems of service chain segmentation and sub-service chain mapping.Moreover,in order to solve the computational complexity problem encountered by the mathematical optimization method in large-scale networks,we model the service function chain based on hidden markov model,and proposes a meta-heuristic algorithm based on an improved Viterbi method.Finally,a large number of comparative simulations are carried out to analyze and compare the performance of our proposed algorithm in networks of different sizes.(2)To solve the reliability problem of service chain in SINET,we propose a reliability-aware service chain resource adapatation mechanism.First,we design the reliability model of the service function chain.On this basis,we apply the redundancy strategy to satisfy the different reliability requirements of various network services.And the reliability-guaranteed service chain mapping problem is modeld as an optimization problem based on mixed integer linear programming with the goal of minimizing the additional resource cost generated by the redundant service chain.Then,we propose a reliability-aware service chain mapping algorithm.It improves the artificial bee colony algorithm by introducing the the penalty function and Deb criterion to solve the constrained optimization problem.The simulation results show that the proper parameter setting is very important for the performance of the proposed algorithm,and it is verified that the redundancy strategy is feasible and effective for improving the reliability of the service chain.In addition,by the comparison simulations,the proposed algorithm has advantages in terms of the satisfaction rate of service requests and the average cost.(3)To realize better enengy efficiency in SINET,we propose a energy-efficient service chain resource adaptation mechanism.Firstly,we present the energy consumption model of service chain.Then,considering the constraints for service chain request,we propose an energy-aware service chain mapping algorithm for calculating the service function path with the lowest power consumption,to order to improve the energy efficiency of the infrastructure network.In addition,with the service chain requests are arriving and expiring continually,the resource distribution of the infrastructure network will be changed dynamically.Hence,we propose an energy-aware service chain migration algorithm,which could update the optimal service function paths for the running service chains by resource integration.After extensive simulations,it shows that the proposed algorithm has good performance in terms of the average value and the real-time value of power dissipation with a low blocking rate.