Integral-Valued Pythagorean Fuzzy-Set-Based Dyna Q+ Framework for Task Scheduling in Cloud Computing

Area of Science:

• Computer Science
• Artificial Intelligence
• Cloud Computing

Background:

• Task scheduling in cloud computing is NP-Hard, complicated by uncertainties in security, traffic, workload, availability, and price.
• Accurate measurement of these uncertainty parameters is challenging, as empirical data often differs from real-world values.
• Existing methods struggle to effectively manage these dynamic uncertainties in task scheduling.

Purpose of the Study:

• To develop an intelligent task scheduler that addresses uncertainties in cloud computing environments.
• To integrate the integral-valued Pythagorean fuzzy set (IVPFS) framework with the Dyna Q+ algorithm for enhanced decision-making.
• To improve the performance metrics of cloud task scheduling, including execution time, makespan, cost, and resource utilization.

Main Methods:

• The Dyna Q+ algorithm, an enhancement of the Dyna Q agent for dynamic environments, was utilized.
• The integral-valued Pythagorean fuzzy set (IVPFS) mathematical framework was incorporated into the Dyna Q+ agent to handle parametric uncertainties.
• The proposed IVPFS Dyna Q+ task scheduler was simulated and evaluated using the CloudSim 3.3 simulator.

Main Results:

• The IVPFS Dyna Q+ task scheduler demonstrated significant performance improvements.
• Execution time and makespan time were reduced by 90%.
• Operation costs were reduced by over 50%, and resource utilization rate improved by 95%.
• Expected value analysis validated the scheduler's effectiveness.

Conclusions:

• The integration of IVPFS with the Dyna Q+ algorithm provides an effective solution for intelligent task scheduling under uncertainty.
• The proposed method achieves a superior balance between exploration and exploitation through action-based learning.
• The results confirm the practical viability and efficiency of the developed task scheduler for cloud computing systems.