A binary Bird Swarm Optimization technique for cloud computing task scheduling and load balancing

• Published in: 2022 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES) pp. 1 - 6
• Main Authors: B, Magesh Kumar, Kumar, M Sathish, Shadrach, Finney Daniel, Polamuri, Subba Rao, R, Poonkodi, Pudi, Vasudeva Naidu
• Format: Conference Proceeding
• Language: English
• Published: IEEE 15.07.2022
• Subjects: Artificial neural networks; Birds; Cloud computing; CNN; Load management; Metaheuristics; neural network; Processor scheduling; Virtual machining
• DOI: 10.1109/ICSES55317.2022.9914085

Cloud computing is a new paradigm for highperformance computation that combines a wide collection of autonomous mixed devices with a flexible computational construction to manage and deliver services. High-performance cloud computation is a subset of cloud computation. When it derives to improving the overall performance of cloud computation, task preparation is one of the most difficult things to improve. This can include response time, make span, and the degree of imbalance. When it comes to reducing power usage, processing time and increasing profit for service providers, task scheduling is vital. Heuristic methods like Binary bird swarm optimization (BBSO) were created to address the problem of inefficient procedures. However, if these algorithms are not paired with additional experiential or meta-heuristic procedures, the optimal solution will not be produced. Because of their high temporal complexity, these algorithms are less useful in real-world scenarios. BSO's binary form is being proposed for cloud computing workload scheduling and balancing in the NP-problem. Our goal function determines if heterogeneous VMs have the greatest completion time difference by taking into account the updating and optimization limits discussed in this research. We developed a technique for updating particle positions in conjunction with load balancing. Metaheuristics and heuristics fail to outperform the proposed technique when it comes to job scheduling and load balancing. Achieved this level of success thanks to the deployment of an artificial neural network. In terms of resource allocation, ANN has shown encouraging outcomes. CNNs are more accurate and faster in predicting targets than multilayer perceptron networks.