AdPSO: Adaptive PSO-Based Task Scheduling Approach for Cloud Computing.

Area of Science:

• Cloud Computing
• Artificial Intelligence
• Operations Research

Background:

• Cloud computing is a preferred platform for research and industry.
• Load-balanced task scheduling is a critical challenge in cloud environments.
• Swarm intelligence algorithms offer promising solutions for cloud scheduling and load balancing.

Purpose of the Study:

• To develop an adaptive Particle Swarm Optimisation (PSO) based task scheduling approach.
• To introduce a novel adaptive inertia weight strategy, Linearly Descending and Adaptive Inertia Weight (LDAIW).
• To enhance task execution time, throughput, and Average Resource Utilization Ratio (ARUR) in cloud environments.

Main Methods:

• Implemented an adaptive PSO algorithm for task scheduling.
• Introduced the LDAIW strategy to balance local and global search in PSO.
• Evaluated the proposed approach against existing PSO inertia weight strategies and other meta-heuristic algorithms.

Main Results:

• The proposed LDAIW strategy achieved significant improvements in makespan (up to 10%), throughput (up to 12%), and ARUR (up to 60%).
• Demonstrated a superior balance between local and global search for optimized task scheduling.
• Outperformed five renowned PSO-based inertia weight strategies and four other meta-heuristic approaches.

Conclusions:

• The adaptive PSO approach with LDAIW is effective for optimizing cloud task scheduling.
• The proposed method offers a practical solution for improving cloud resource management and performance.
• Further research can explore variations of the LDAIW strategy for diverse cloud workloads.