CAJ | 학술논문

对私有云计算平台上资源最优路径匹配,实现对资源的有效调度和检测,提高资源共享能力.传统算法采用云平台的资源置换算法实现云资源目标匹配,受限于机器学习技术自身的复杂性,性能不好.提出一种基于委托管理节点角色量化合成的私有云计算平台上资源最优搜索路径匹配算法.构建私有云计算平台终端数据访问和资源调度模型,根据贪心算法的收缩原则,在资源访问管理中引入互斥锁机制以保证资源共享操作的完整性,路径匹配有向图模型,提高资源路径匹配搜索能力.以全局度量为中心节点,计算逆向追踪搜索频度,实现算法优化.仿真实验结果表明,采用该算法进行私有云计算平台上资源最优路径匹配优化,能有效提高私有云平台上的资源路径匹配准确度,从而提高资源搜索成功率,实现资源优化共享,展示优越的适用价值.
대사유운계산평태상자원최우로경필배,실현대자원적유효조도화검측,제고자원공향능력.전통산법채용운평태적자원치환산법실현운자원목표필배,수한우궤기학습기술자신적복잡성,성능불호.제출일충기우위탁관리절점각색양화합성적사유운계산평태상자원최우수색로경필배산법.구건사유운계산평태종단수거방문화자원조도모형,근거탐심산법적수축원칙,재자원방문관리중인입호척쇄궤제이보증자원공향조작적완정성,로경필배유향도모형,제고자원로경필배수색능력.이전국도량위중심절점,계산역향추종수색빈도,실현산법우화.방진실험결과표명,채용해산법진행사유운계산평태상자원최우로경필배우화,능유효제고사유운평태상적자원로경필배준학도,종이제고자원수색성공솔,실현자원우화공향,전시우월적괄용개치.
The private cloud computing platform resources, optimal path, realize the effective scheduling and detection of re-sources, improve resource sharing ability. The traditional algorithm uses the resource substitution algorithm for cloud plat-form, achieve the goal of cloud resources, limited by the complexity of the machine learning technology itself, the perfor-mance is not good. Presented a method for calculating the optimal path algorithm of matching resources platform commis-sioned management node role quantitative synthesis based on private cloud. Building a private cloud computing platform terminal access to data and resources scheduling model, according to the contraction principle of greedy algorithm, in the resource access and management in the introduction of mutex lock mechanism to guarantee the resources sharing operation-al integrity, path matching the digraph model, improve resource path matching search ability. With the global metric as the center node, computing reverse tracing search frequency, algorithm optimization. Simulation results show that, by using this algorithm to a private cloud computing platform resources, optimal path optimization, can effectively improve the resource path private cloud platform on the matching accuracy, so as to improve the success rate of resource searching, realize the op-timization of resource sharing, show superior application value.