| 研究実績の概要 |
First, I have developed a data-placement technique for data-center resizing that considers both the static and dynamic characteristics of data centers. The objective is to find an optimal data placement to ameliorate the performance degradation when turning off idle nodes, and to improve the performance of big-data processing when turning on more working nodes by optimizing global data transfer cost.
Second,I have developed a data-placement algorithm for big-data processing that considers network traffic reduction as well as QoS guarantees for the data blocks. Our objective is to optimize data placement so as to minimize data transfers and guarantee the QoS requirements to improve the performance of big-data processing such that the overall computation, communication cost, and isolated data blocks are minimized.
Third, I have developed streaming workflow allocation algorithms to help cloud infrastructure better support the real-time data analysis with minimal cost by: 1) optimizing the specification of streaming workflow; 2) using MCMF algorithm to assign streaming workflow into geo-distributed data centers with star network topology; 3) using LP algorithm with two computational space reduction methods to assign streaming workflow into geo-distributed data centers with general network topology.
|
