Scholarlink

Background <p>Cloud computing offers on-demand services requires efficient resource utilization to ensure the quality and performance of the services. This necessitates intelligent mechanisms for effective resource management in autonomic cloud environments.</p> Problem Definition <p>However, accurate prediction on future resource requirements is a challenging task due to varying workloads and the heterogeneous nature of cloud infrastructures.</p> Method description <p>To address the mentioned issue, here proposed an Enhanced Resource Provisioning Prediction in Autonomic Cloud Computing Using Optimized Quaternion Generative Adversarial Networks (RPP-ACCE-QGAN). Initially PlanetLab dataset is utilized for data acquisition and preprocessing is processed using Multi-observation Fusion Kalman Filter (MOFKF) to handle missing values and enhance data quality. Then, the preprocessed data are fed into the Quaternion Generative Adversarial Networks (QGAN) for resource usage prediction. The Bitwise Arithmetic Optimization Algorithm (BAOA) is employed to fine-tune the hyperparameters of QGAN model.</p> Results <p>Experimental results exemplify that the proposed RPP-ACCE-QGAN model attains enhanced prediction performance compared with existing methods, such as RUP-CCE-DCRNN, AAS-CRP-CNN and DRP-CCE-LSTM. The proposed framework demonstrates significant enhancement in terms of 23.20%, 21.15%, and 16.19% reduced latency and 23.20%, 21.15%, and 16.19% reduced error metrics including MAE, RMSE and MAPE.</p> Conclusion <p>Overall, the proposed RPP-ACCE-QGAN model attains an effective solution for resource provisioning prediction in autonomic cloud computing environments.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhanced resource provisioning prediction in autonomic cloud computing using optimized quaternion generative adversarial networks

  • 期刊论文

  • 发表日期：

    2026年5月29日

查看全文
  • T Karpagam,
  • A Syed Ismail,
  • K Suresh Babu,
  • K Krishnamoorthi

摘要

Background

Cloud computing offers on-demand services requires efficient resource utilization to ensure the quality and performance of the services. This necessitates intelligent mechanisms for effective resource management in autonomic cloud environments.

Problem Definition

However, accurate prediction on future resource requirements is a challenging task due to varying workloads and the heterogeneous nature of cloud infrastructures.

Method description

To address the mentioned issue, here proposed an Enhanced Resource Provisioning Prediction in Autonomic Cloud Computing Using Optimized Quaternion Generative Adversarial Networks (RPP-ACCE-QGAN). Initially PlanetLab dataset is utilized for data acquisition and preprocessing is processed using Multi-observation Fusion Kalman Filter (MOFKF) to handle missing values and enhance data quality. Then, the preprocessed data are fed into the Quaternion Generative Adversarial Networks (QGAN) for resource usage prediction. The Bitwise Arithmetic Optimization Algorithm (BAOA) is employed to fine-tune the hyperparameters of QGAN model.

Results

Experimental results exemplify that the proposed RPP-ACCE-QGAN model attains enhanced prediction performance compared with existing methods, such as RUP-CCE-DCRNN, AAS-CRP-CNN and DRP-CCE-LSTM. The proposed framework demonstrates significant enhancement in terms of 23.20%, 21.15%, and 16.19% reduced latency and 23.20%, 21.15%, and 16.19% reduced error metrics including MAE, RMSE and MAPE.

Conclusion

Overall, the proposed RPP-ACCE-QGAN model attains an effective solution for resource provisioning prediction in autonomic cloud computing environments.