为满足车联网中海量数据的采集、传输以及对这些数据的快速处理的需求,可采用移动边缘计算(MEC)技术.本文考虑移动边缘计算中基站连接方式和物理资源的特点,对边缘服务器的部署问题进行了分析,以部署成本和网络时延为优化目标,划分基站集群,并使用整数线性规划(ILP)建立模型.为了获得运行效率更高的边缘服务器部署方案,本文使用分支定界算法和启发式贪婪算法获得优化模型的近似最优解.实验评估结果显示,分支定界算法和启发式贪婪算法最高可以把边缘服务器部署算法运行时间减少37.6％.此外,本文分析了用户服务器请求数量和用户服务优先级对算法运行时间和边缘服务器运行成本的影响.

More and more mobile devices and applications are beginning to use undirected networks. In order to improve the user experience and ensure the quality of network services, edge computing is proposed, which is also an important part of the fifth generation network. As a key technology of 5G networks, NFV and SDN are receiving more and more attention. NFV uses virtualization technology to flexibly deploy network services. At the same time, SDN can provide all information about network topology and resources. The combination of NFV and SDN provides a new approach to network service deployment in edge computing. Due to the performance limitations of the edge server, when the users and devices accessing the network increase, if the edge server cannot provide the corresponding network resources, the network service quality will be degraded. This paper proposes a dynamic virtual network service placement scheme based on network service type, summarizes the characteristics of different network service types for network resource usage, determines the service type when users request network services, and proposes to ensure the optimal utilization of network resources and network delay. The algorithm is dynamically placed and the feasibility of the proposed scheme is proved by experiments.

Network Function Virtualization (NFV) is a rapidly evolving network technology in recent years. The purpose of NFV is to use virtualization technology to softwareize network functions, and dynamically deploy virtual network functions (VNFs) according to the usage status of network links and the service requirements of users. NFV can increase the flexibility of network services and the utilization of network resources. In the proposed paper, we analyze the user data of urban computing, and propose that the time and location of the user’s use of the network service is subject to regular changes. Based on this judement, we propose a new energy-saving deployment method for virtual data centers (vDCs). In this paper, we formalize the placement problem of vDC into a multicommodity flow problem and address it as an integer linear programming (ILP). We design a centrality-based greedy algorithm and evaluate its effectiveness by comparing the proposed algorithm with the ILP optimal solution. The evaluation results show that the greedy algorithm proposed in this paper can obtain the approximate optimal solution of ILP, and the running time of the proposed algorithm is shorter than the ILP solution when the number of network nodes increases.

SDN traffic engineering plays an important role in today’s Internet, cloud computing and sensor network. Source routing combined with SDN traffic engineering has broad application prospects. There are two processing procedures of SDN traffic engineering when building a model, the difference of them is whether to list all candidate paths firstly or not. In order to make the traffic engineering achieves the best results, for a source-destination pair, the form of the solution of general SDN traffic engineering model is the load on each link in the network. However, based on this form of the solution, we cannot conFigure source routing directly. In this paper, we propose the problem of path allocation based on source routing that helps to deploy source routing configuration at the source node. We model and analyze this problem, then a relatively simplified method that obtains forwarding paths sequentially is proposed. The evaluation results indicate that the proposed method can address this problem effectively, and can meet traffic demands as much as possible by using source routing. In addition, the favorable effects of the combination of source routing and SDN are also verified.

In order to address the multi-commodity flow problem for traffic scheduling in software-defined networking,a method based on segment routing was proposed.The proposed method pre-computed sets of candidate paths and attributes of these paths for all source-target nodes,and set the requirements of attributes of candidate paths that should be met combined with various demands and constraints of flows,then generated sets of candidate paths for flows.In the proposed scheme,multi-commodity flow model in software-defined networking was simplified based on sets of candidate paths for flows,the difficulty of solving was reduced,the centralized control by the controller and the autonomous control by nodes were supported,the scalability of controller was improved.In addition,how to meet the energy-saving needs of the network was proposed,i.e.,reducing the number of links that could participate in flow forwarding.The performance evaluation results indicate that the proposed method can meet various demands and constraints of flows,improve network performance,and reduce the computational load of solving the problem of traffic scheduling.

网络功能虚拟化(NFV)是视频流应用的重要技术.在视频流使用场景下,已有的工作研究NFV网络中的虚拟深度包检测(vDPI)放置问题时,仅考虑减少vDPI放置的数量,但没有考虑放置vDPI功能带来的NFV网络性能稳定性问题.针对上述不足,本文在减小vDPI放置数量的基础上,考虑提高网络稳定性,提出了一个多目标线性整数规划模型,并设计了一种贪心近似放置算法.该算法在NFV中放置vDPI功能时,可以降低vDPI放置数量、减少流量经过的平均网络跳数、降低传输时延、保证网络性能的稳定性.本文采用Lingo求出数学模型的最优解,并用贪心近似放置算法进行实验,对比算法实验结果与模型所求最优解可知:所提算法正确性较高,有较好的时间复杂度,适用于不同规模的NFV网络.