嵌入式网络信息接口数据实时采集仿真

摘要

对嵌入式网络信息接口的数据实时采集,可以有效解决网络信息内容杂乱无序的问题.对接口数据进行实时采集,需要先调整各个数据变量间的相关性,组建数据变量间的协方差矩阵,完成接口数据的实时采集.传统方法利用采集样本,计算到聚类中心模糊加权距离总和,但无法得到数据变量间的协方差矩阵,导致采集精度低.提出基于多元时序模糊聚类分段的网络信息整合下嵌入式网络信息接口数据实时采集方法.在给定的连续条件下对接口全部数据依据相似性进行聚类处理,组建数据变量间的协方差矩阵,依据接口数据的变化贡献率提取其特征值,选取相应的特征向量作为数据空间中的坐标轴,融合于匈牙利方法计算各个特征坐标系间的做小距离,并完成对嵌入式网络信息接口数据实时采集.实验结果表明,所提方法采集精确度高,为保障网络的服务质量提供了科学的依据.%A real-time collection method for data of information interface in embedded network under network information integration is proposed based on fuzzy clustering segmentation of multivariate time series.Firstly,all data of interface under given continuity condition are clustered according to similarity,and covariance matrix among data variables is built.The feature value of interface data is extracted according to its rate of variation contribution.Then,the corresponding feature vector is selected as coordinate axis in data space.Finally,the minimum distance between each feature coordinate system integrated with the Hungarian method is calculated and the real-time collection is completed.Simulation results show that the proposed method has high collection precision.It provides scientific gist for ensuring quality of network service.