Smart load balancing in cloud computing: Integrating feature selection with advanced deep learning models.

Area of Science:

• Cloud Computing
• Artificial Intelligence
• Resource Management

Background:

• Cloud computing's growth presents resource management challenges.
• Traditional load balancing is insufficient for dynamic cloud environments.
• Suboptimal utilization and high costs result from inadequate load balancing.

Purpose of the Study:

• To introduce a novel smart load-balancing strategy for cloud environments.
• To address the limitations of conventional methods in handling dynamic workloads.
• To enhance resource utilization and reduce operational costs in cloud infrastructures.

Main Methods:

• Proposed the Smart Load Adaptive Distribution with Reinforcement and Optimization (SLADRO) approach.
• Integrated Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) for load prediction.
• Utilized Orthogonal Arrays and Particle Swarm Optimization (OOA-PSO) for feature selection and Deep Reinforcement Learning (DRL) for task scheduling.

Main Results:

• SLADRO significantly outperforms traditional load-balancing techniques.
• Demonstrated notable improvements in throughput and makespan.
• Achieved enhanced resource utilization and energy efficiency.

Conclusions:

• SLADRO offers a scalable and adaptive solution for cloud load balancing.
• The hybrid methodology effectively optimizes resource allocation.
• Advanced techniques provide a comprehensive framework for efficient cloud resource management.