Deadline-aware and energy-efficient IoT task scheduling in fog computing systems: A semi-greedy approach

Journal of Network and Computer Applications, Journal Year: 2022, Volume and Issue: 201, P. 103333 - 103333

Published: Jan. 26, 2022

Language: Английский

---

Sustainable computing across datacenters: A review of enabling models and techniques

Computer Science Review, Journal Year: 2024, Volume and Issue: 52, P. 100620 - 100620

Published: Feb. 13, 2024

Language: Английский

---

A Systematic Review of Energy Management Strategies for Resource Allocation in the Cloud: Clustering, Optimization and Machine Learning

Energies, Journal Year: 2021, Volume and Issue: 14(17), P. 5322 - 5322

Published: Aug. 27, 2021

Nowadays, many organizations and individual users are employing cloud services extensively due to their efficiency, reliability low cost. A key aspect for data centers is achieve management methods reduce energy consumption, increasing the profit reducing environmental impact, which critical in deployment of leading-edge technologies today such as blockchain digital finances, IoT, online gaming video streaming. In this review, various clustering, optimization, machine learning used resource allocation increase efficiency performance analyzed, compared classified. Specifically, on one hand, we discuss how clustering optimization techniques widely applied capacity provide solutions consumption reduction. On other study multi-objective focus well service level agreement (SLA) violation, improving quality (QoS) simultaneously. Also, firefly algorithm, whale algorithm (WOA), particle swarm (PSO) genetic (GA) highest field. Moreover, analyze deep neural network (DNN), random forest, support vector (SVM) prediction cloud, showing an accurate prediction. Nevertheless, existing still have limitations convergence, trap into local optima overfitting.

Language: Английский

---

GRVMP: A Greedy Randomized Algorithm for Virtual Machine Placement in Cloud Data Centers

IEEE Systems Journal, Journal Year: 2020, Volume and Issue: 15(2), P. 2571 - 2582

Published: June 30, 2020

Cloud computing efficiency greatly depends on the of virtual machines (VMs) placement strategy used. However, VM has remained one major challenging issues in cloud mainly because heterogeneity both and physical (PMs), multidimensionality resources, increasing scale data centers (CDCs). An inefficiency a significant influence quality service provided, amount energy consumed, running costs CDCs. To address these issues, this article, we propose greedy randomized (GRVMP) algorithm large-scale CDC with heterogeneous multidimensional resources. GRVMP inspires "power two choices" model places VMs more power-efficient PMs to jointly optimize usage resource utilization. The performance is evaluated using synthetic real-world production scenarios (Amazon EC2) several matrices. results experiment confirm that optimizes power overall wastage also show significantly outperforms baseline schemes terms metrics

Language: Английский

---

An efficient power-aware VM allocation mechanism in cloud data centers: a micro genetic-based approach

Cluster Computing, Journal Year: 2020, Volume and Issue: 24(2), P. 919 - 934

Published: Aug. 9, 2020

Language: Английский

---

A priority, power and traffic-aware virtual machine placement of IoT applications in cloud data centers

Journal of Systems Architecture, Journal Year: 2021, Volume and Issue: 115, P. 101996 - 101996

Published: Jan. 14, 2021

Language: Английский

---

Power efficient virtual machine placement in cloud data centers with a discrete and chaotic hybrid optimization algorithm

Cluster Computing, Journal Year: 2020, Volume and Issue: 24(2), P. 1293 - 1315

Published: Sept. 28, 2020

Language: Английский

---

Toward an autonomic approach for Internet of Things service placement using gray wolf optimization in the fog computing environment

Software Practice and Experience, Journal Year: 2021, Volume and Issue: 51(8), P. 1745 - 1772

Published: May 16, 2021

Abstract Divers and the huge amount of data produced by Internet Things (IoT) applications on one hand, inherent limitations local equipment to handle these data, other leads present emerging closer technologies end‐users such as fog computing environment. Nevertheless, despite numerous advantages an environment, it still needs state‐of‐the‐art approaches cope with some limitations. In literature, resource placement strategies are generally proposed address problems, in which IoT mapped nodes. However, its importance, different attempt enhance overall system's performance users' expectations: none is satisfactory. this article, deploy nodes, autonomic service approach based gray wolf optimization scheme proposed, enhancing while considering execution costs. Besides, concepts help make appropriate automanagement system that fits better environment's dynamic behavior. Simulation results demonstrate outperforms converges solution near‐optimal application deployment nodes respect performing services 93.7%, average waiting time for performed 100%, remaining sent extra provisioned period zero.

Language: Английский

---

A Predictive Priority-Based Dynamic Resource Provisioning Scheme With Load Balancing in Heterogeneous Cloud Computing

IEEE Access, Journal Year: 2021, Volume and Issue: 9, P. 62653 - 62664

Published: Jan. 1, 2021

In cloud computing, resource provisioning is a key challenging task due to dynamic for the applications. As per workload requirements of application's resources should be dynamically allocated application. Disparities in produce energy, cost wastages, and additionally, it affects Quality Service (QoS) increases Level Agreement (SLA) violations. So, applications quantity match with required quantity. Load balancing computing can addressed through optimal scheduling techniques, whereas this solution belongs NP-Complete optimization problem category. However, providers always face management issues variable workloads heterogeneous system environment. This issue has been solved by proposed Predictive Priority-based Modified Heterogeneous Earliest Finish Time (PMHEFT) algorithm, which estimate upcoming demands. research contributes towards developing prediction-based model efficient heterogamous environment fulfill end user's requirements. Existing algorithms fail meet such as makespan minimization budget constraints satisfaction, or incorporate principles, i.e., elasticity heterogeneity resources. paper, we PMHEFT algorithm minimize given workflow application improving load across all virtual machines. Experimental results show that our algorithm's makespan, efficiency, power consumption are better than other algorithms.

Language: Английский

---

Workload time series prediction in storage systems: a deep learning based approach

Cluster Computing, Journal Year: 2021, Volume and Issue: 26(1), P. 25 - 35

Published: Jan. 13, 2021

Language: Английский

---