閱讀全文 | |
篇名 |
Task Scheduling and Resource Allocation Based on Ant-Colony Optimization and Deep Reinforcement Learning
|
---|---|
並列篇名 | Task Scheduling and Resource Allocation Based on Ant-Colony Optimization and Deep Reinforcement Learning |
作者 | Ulysse Rugwiro、Chunhua Gu、Weichao Ding |
英文摘要 | Cloud computing has become a significant aspect of today’s rapidly growing technology, accessing as it does a large number of servers, given users’ constant need to access their data efficiently and quickly. Cloud computing providers can flexibly place a user’s task into an appropriate virtual machine and allocate the resource to the tasks for proper execution. However, user tasks can take a long time to complete the execution when the required resources are not available on the server. To overcome this problem, we propose a task scheduling and resource allocation model based on Hybrid Ant Colony Optimization and Deep Reinforcement Learning. In this article, our goal is to minimize the overall task completion time and improve the utilization of idle resources. The task scheduling was performed by constructing a Binary In-order Traversal Tree using weighted values. We then introduced a Deep Reinforcement Learning (DRL) algorithm to reduce space complexity by splitting resources into state space and action space. A state space will contain idle resources, which are used in task allocation. Then the scheduled task will search the resources based on Ant Colony Optimization. When it finds an optimal resource, it will allocate it to the task, and the server will put the allocated resources into action space. If the VM is overloaded, migration is performed. We simulated the proposed algorithm using CloudSim and evaluated the performance in terms of task completion time and resource utilization. Our proposed work evaluation shows mitigation of the above-described problems and illustrates the reduction of waiting time and improvement in idle resource utilization. |
起訖頁 | 1463-1475 |
關鍵詞 | Task scheduling、Resource allocation、Ant Colony Optimization、Deep Reinforcement Algorithm |
刊名 | 網際網路技術學刊 |
期數 | 201909 (20:5期) |
出版單位 | 台灣學術網路管理委員會 |
DOI |
|
QR Code | |
該期刊 上一篇
| Dynamic Optimization Algorithm of Static Materialized Views |
該期刊 下一篇
| Peer-assisted Data Offloading and Distributed Channel Selection for Mobile Cloud Computing |