ID 原文 译文
44636 为此，提出了一种基于 P4 的能够覆盖全网且可扩展性强的主动网络遥测平台 NetVision。 For this reason, a proactive net-work telemetry platform NetVision was proposed based on P4 that can cover the whole network and has good scalability.
44637 随着互联网的高速发展和移动终端设备的快速普及，传统服务流程向高实时、高可靠性的移动式服务组合方向发展。 With the development of internet and the popularization of mobile equipment, the traditional service process has been developing towards to the direction of high real-time and reliability.
44638 但由于网络地址呈爆发式增长，亟需 IPv6 技术来解决 IPv4 面临的地址短缺的问题。 However, due to the explosive growth of network addresses, IPv6 technology is urgently needed to solve the current shortage problem of IPv4 addresses.
44639 考虑到移动网络的服务资源位置变动性、网络连接的高频率不稳定性、服务资源的受限性会直接影响服务流程质量，研究支持移动环境的服务流程不确定性评估推荐方法，尤其是在 IPv6 环境下部署服务和组合服务，提出面向数据感知的移动服务流程服务质量描述方法。 Considering the position variability, high frequency instability, and service resource constrained, the recommendation method focusing on the uncertainty evaluation, especially deploying services and composting services under the environment ofIPv6 environment were studied, data-aware oriented mobile service process and QoS description method were proposed.
44640 首先，从数据感知角度对服务流程的数据进行建模，定义不同结构的服务流程模型的非功能性性质计算方法。 First, the data model of service process was formalized, defining the non-functional properties for different structures of service process.
44641 其次，基于皮尔逊相关系数计算用户与用户之间的关联度，改进天际线方法对服务流程推荐结果进行优选。 Second, the relationship between users and users is calculated based on Pearson correlation coefficient, and the recommendation results of service process are optimized by improving the Skyline method.
44642 再次，基于用户评分设置非功能性参数，提出服务流程非功能性归一化计算公式，从而实现推荐 Top-k 个服务流程。 Third, the non-functional para-meters are set according to user ratings, and the normalization calculation formula for service process is proposed to return Top-k service processes.
44643 最后，结合实验分析证明该方法的可行性。 Finally, the results of a series of experiments show that the method is feasible.
44644 为了解决校园网络日志存储中流量变化导致的机械硬盘平均访问时间长、波动大的问题，设计了一种通过流量预测进行日志写入的方法。 In order to solve the problem of long average access time and large fluctuation of hard disk caused by traffic-changes in campus network log storage, a method with log writing through traffic prediction was designed.
44645 首先，提出流量定界算法对流量进行划分为高峰期、中间期与低谷期。 First, a traffic partition algorithm to divide traffic to three periods was proposed：peak period, intermediate period and low period.