文献类型：期刊文献

中文题名：分布式网络资源用户信息快速获取仿真

英文题名：Simulation of Rapid Acquisition of User Information in Distributed Network Resources

作者：暴占彪[1]

第一作者：暴占彪

机构：[1]河南财经政法大学网络信息管理中心

第一机构：河南财经政法大学

年份：2019

卷号：0

期号：4

起止页码：345-348

中文期刊名：计算机仿真

外文期刊名：Computer Simulation

收录：CSTPCD;;北大核心:【北大核心2017】；

语种：中文

中文关键词：分布式网络;网络资源;用户信息;快速获取

外文关键词：Distributed network;Network resource;User information;Fast acquisition

摘要：对分布式网络资源用户的信息进行快速获取,能够有效整合网络用户,提高网络运行效率。对资源用户信息的获取,需要对BP神经网络误差函数权值进行更新调整,对数据结构神经元之间连接权值进行修正。传统方法根据本征模函数上的数据结构包络,采用3α原则与计算获取阈值的对比,但忽略了对数据结构权值的修正,导致获取精度偏低。提出一种基于BP神经网络的资源用户信息快速获取方法,对数据结构原信号的尺度缺失进行补充,采用BP神经网络计算其误差函数,同时利用梯度下降法对BP神经网络误差函数权值进行更新调整;引入动量因子α,将BP神经网络学习模式不断输入到输入层,对数据结构神经元之间的连接权值进行修正,实现对分布式网络资源用户信息的快速获取。实验结果显示,所提方法具有高快速获取率、低误报率的资源用户信息快速获取性能,且大大提高了收敛速度。
The fast acquisition of user information in distributed network resources can effectively integrate network users and improve running efficiency.The traditional method compares the 3αprinciple with the threshold value,but ignores the correction of weight value of data structure,resulting in the low acquisition precision.Therefore,this article focuses on a method to fast acquire user information based on BP neural network.This research supplied the scale deficiency of the original signal of data structure and used BP neural network to calculate its error function.Then,our research used the gradient descent method to update and adjust the weight value of error function of BP neural network.Moreover,the research introduced the momentum factorαto constantly input the learning mode of BP neural network into the input layer.Finally,we modified the connection weight value between data structure neurons to achieve the rapid acquisition of user information in distributed network resource.Simulation results show that the proposed method has fast acquisition performance of resource user information with fast acquisition rate and low false alarm rate,which greatly improves the convergence rate.