基于深度学习的异构资源分配算法研究  

Research on heterogeneous resource allocation algorithms based on deep learning

作　　者：冯硕 杨军[1] 张鹏飞[1] FENG Shuo;YANG Jun;ZHANG Peng-fei(North China Institute of Computing Technology,Beijing 100083,China)

机构地区：[1]华北计算技术研究所基础四部

出　　处：《信息技术》2020年第1期116-120,共5页Information Technology

摘　　要：资源分配是目前云计算领域中一个重要的研究方向。在异构云计算体系结构下的复杂应用问题研究中,为了满足异构资源分配的需求,提升资源利用效率,文中提出了一种基于深度学习的面向应用的资源分配算法。该算法将数据特征进行量化,更加精确地刻画了不同服务器资源之间的性能差异,在分配算法中加入了一个工作负载预测模型,使给出的资源分配方案与需求更加匹配,同时提高了资源利用率。Resource allocation is an important research direction in the field of cloud computing.In the study of complex application problems under heterogeneous cloud computing architecture,in order to meet the needs of heterogeneous resource allocation and improve resource utilization efficiency,an application-oriented resource allocation algorithm based on deep learning is proposed.The algorithm quantifies the data features,more accurately characterizes the performance difference between different server resources,and adds a workload prediction model to the allocation algorithm,which makes the given resource allocation scheme more compatible with the requirements,and improves resource utilization rate.

关 键 词：资源分配 异构云 神经网络 资源利用率 

分 类 号：TP393.0[自动化与计算机技术—计算机应用技术]