期刊名称：International Journal of Electrical and Computer Engineering
电子版ISSN：2088-8708
出版年度：2018
卷号：8
期号：3
页码：1629-1635
DOI：10.11591/ijece.v8i3.pp1629-1635
语种：English
出版社：Institute of Advanced Engineering and Science (IAES)
摘要：In this article, a searching method for the rational task distribution through the nodes of a hyperconverged network is presented in which it provides the rational distribution of task sets towards a better performance. With using new subsettings related to distribution of nodes in the network based on distributed processing, we can minimize average packet delay. The distribution quality is provided with using a special objective function considering the penalties in the case of having delays. This process is considered in order to create the balanced delivery systems. The initial redistribution is determined based on the minimum penalty. After performing a cycle (iteration) of redistribution in order to have the appropriate task distribution, a potential system is formed for functional optimization. In each cycle of the redistribution, a rule for optimizing contour search is used. Thus, the obtained task distribution, including the appeared failure and success, will be rational and can decrease the average packet delay in the hyperconverged networks. The effectiveness of our proposed method is evaluated by using the model of hyperconverged support system of the university E-learning provided by V. N. Karazin Kharkiv National University. The simulation results based on the model clearly confirm the acceptable and better performance of our approach in comparison to the classical approach of task distribution.
其他摘要：In this article, a searching method for the rational task distribution through the nodes of a hyperconverged network is presented in which it provides the rational distribution of task sets towards a better performance. With using new subsettings related to distribution of nodes in the network based on distributed processing, we can minimize average packet delay. The distribution quality is provided with using a special objective function considering the penalties in the case of having delays. This process is considered in order to create the balanced delivery systems. The initial redistribution is determined based on the minimum penalty. After performing a cycle (iteration) of redistribution in order to have the appropriate task distribution, a potential system is formed for functional optimization. In each cycle of the redistribution, a rule for optimizing contour search is used. Thus, the obtained task distribution, including the appeared failure and success, will be rational and can decrease the average packet delay in the hyperconverged networks. The effectiveness of our proposed method is evaluated by using the model of hyperconverged support system of the university E-learning provided by V. N. Karazin Kharkiv National University. The simulation results based on the model clearly confirm the acceptable and better performance of our approach in comparison to the classical approach of task distribution.
