可信云计算平台多源大数据时间序列调度算法

Trusted cloud computing platform poly source big data time sequence scheduling algorithm

可信云计算平台多源大数据时间序列调度性能过差会增加平台传输能耗和运营成本,降低多源大数据利用率,为使平台内数据能根据任务目标合理完成时间序列调度工作,提出可信云计算平台多源大数据时间序列调度算法,该方法首先构建混沌时间序列模型挖掘可信云计算平台多源大数据,并利用小波阈值降噪方法优化多源大数据,然后将优化后的多源大数据与海量并行贝叶斯因子化分解方法相结合,根据该方法输出的时间序列调度策略,实现可信云计算平台多源大数据时间序列调度。实验结果表明:本文方法加速比最高为97.2%,资源调度总功率仅为2300 kW,负载均衡离差不超过0.2。

The poor performance of time series scheduling for multi-source big data on a trusted cloud computing platform can increase platform transmission energy consumption and operating costs,and decrease the utilization rate of multi-source big data.In order to enable the data within the platform to be reasonably scheduled according to task objectives,a trusted cloud computing platform multi-source big data time series scheduling algorithm is proposed.This method first constructs a chaotic time series model to mine the multi-source big data on the trusted cloud computing platform,and then optimizes the data using the wavelet threshold denoising method.The optimized multi-source big data is then combined with the massive parallel Bayesian factorization decomposition method.Based on the time series scheduling strategy output by this method,the time series scheduling of multi-source big data on a trusted cloud computing platform is realized.Experimental results show that the maximum acceleration ratio achieved b this method is 97.2%,the total power of resource scheduling is only 2300 kW,and the load balance deviation does not exceed 0.2.