基于Java3D网络大数据多分辨率采集方法仿真

针对传统的网络大数据多分辨率采集方法通常存在采集成本高、完成时间较长、采集准确率较低等问题,提出基于Java3D的大数据网络多分辨率采集方法。通过对基于Java3D网络的大数据多分辨率进行时间序列的空间重构,求出多分辨率时间序列的自相关函数,并引入平均互动数据提取方法对网络数据多分辨率采集的功率谱密度进行估计,采用ADASYN算法,对无效多分辨率数据进行去除,实现大数据多分辨率精准采集。实验结果表明,所提方法采集成本较低、完成时间较短、采集准确率较高;具有一定的实用价值,可广泛应用于各个领域。

Simulation of Network Big Data Multi-Resolution Acquisition Method Based on Java3D

In the traditional network big data multi-resolution acquisition,the acquisition cost is high and the completion time is long.Meanwhile,the acquisition accuracy is low.Therefore,this paper proposed a multi-resolution acquisition method for network big data based on Java 3D.On the basis of the spatial reconstruction of time series,the autocorrelation function of multi-resolution time series was obtained.Moreover,the average interactive data extraction method was introduced to estimate the power spectral density of multi-resolution acquisition of network data.Finally,ADASYN algorithm was used to remove the invalid data,and thus to achieve the accurate multi-resolution acquisition of network big data.Simulation results show that the proposed method has low acquisition cost,short completion time and high acquisition accuracy.