Diversity-Aware Marine Predators Algorithm for Task Scheduling in Cloud Computing.

Area of Science:

• Computer Science
• Artificial Intelligence
• Cloud Computing

Background:

• Increasing cloud users and Internet of Things (IoT) applications necessitate advanced task scheduling (TS) methods.
• Efficient resource allocation in cloud computing is crucial for performance and energy conservation.

Purpose of the Study:

• To propose a novel diversity-aware marine predators algorithm (DAMPA) for optimizing task scheduling in cloud computing environments.
• To enhance the performance of task scheduling by addressing premature convergence and balancing exploration-exploitation capabilities.

Main Methods:

• The proposed Diversity-Aware Marine Predators Algorithm (DAMPA) incorporates predator crowding degree ranking and comprehensive learning strategies.
• A stage-independent step-size scaling strategy with varying control parameters is employed to balance exploration and exploitation.
• The algorithm was evaluated through two distinct case experiments.

Main Results:

• DAMPA demonstrated significant reductions in makespan and energy consumption compared to the latest algorithms.
• In the first case, reductions reached up to 21.06% for makespan and 23.47% for energy consumption.
• In the second case, average reductions were 34.35% for makespan and 38.60% for energy consumption, with improved throughput.

Conclusions:

• The DAMPA algorithm effectively addresses task scheduling challenges in cloud computing.
• The proposed methods successfully maintain population diversity, avoid premature convergence, and optimize exploration-exploitation balance.
• DAMPA offers a superior approach for task scheduling, leading to enhanced efficiency and throughput in cloud environments.