项目名称: 基于网络业务特征的用户行为及虚拟映射技术研究
项目编号: No.61201153
项目类型: 青年科学基金项目
立项/批准年度: 2013
项目学科: 电子学与信息系统
项目作者: 崔鸿雁
作者单位: 北京邮电大学
项目金额: 25万元
中文摘要: 针对国家建设未来高可信网络的战略计划需求,本项目针对日益增长的网络数据业务,在研究海量网络数据的特征和用户业务行为需求的基础上,设计未来网络中基于虚拟映射技术的高效资源分配方案,实现未来网络和用户之间通信可信赖的目的,解决目前通信网络缺乏动态预测和资源智能管理能力的问题。具体包括:本项目拟设计一套完整的网络海量数据采集和分析系统;研究海量网络业务数据分类理论;研究网络流量及用户行为预测算法,提出具有动态感知能力的可信网络模型,及研究该模型下基于虚拟映射的资源管理机制。本项目的目标是探索未来网络面临的如何保证和提高网络智能管控能力的关键技术,从而解决当数据业务特征和用户行为需求未知、负载变化、网络节点失效等因素引起的当前网络盲目调度资源问题,为构建一个智能、可信的网络资源管理体系提供理论依据和实践指导。
中文关键词: 用户行为;神经网络;预测;分布式系统;虚拟化
英文摘要: For the needs of building high credibility network, the efficient resource allocation scheme based on virtual mapping is designed for the huge growing data services in the future network to make communication credible between the network and users. It solves the problems of lacking of dynamic prediction and intelligent resources control in the current network.The research work in the project is as following: A mass data collection and analysis system,the data classifacation method, the prediction method of network traffic and user behavior, A credible network architecture with dynamic perception ability,and the resource scheduling mechanism based on virtual mapping in the new network. The goal of the project is how to guarantee and improve the network intelligent control ability, and how to solve the problem of blind network resources scheduling, when the data services characteristics and user behavior requirement, traffics load variation and network nodes failure are unknow. The research work provides theoretical basis and practical guidance to build a smart, trusted network resource management system.
英文关键词: Users Behavior;Neural Network;Prediction;Distributed system;Virtualization