A Secure and Multiobjective Virtual Machine Placement Framework for Cloud Data Center

IEEE Systems Journal, Journal Year: 2021, Volume and Issue: 16(2), P. 3163 - 3174

Published: July 20, 2021

To facilitate cost-effective and elastic computing benefits to the cloud users, energy-efficient secure allocation of virtual machines (VMs) plays a significant role at data centre. The inefficient VM Placement (VMP) sharing common physical among multiple users leads resource wastage, excessive power consumption, increased inter-communication cost security breaches. address aforementioned challenges, novel multi-objective machine placement (SM-VMP) framework is proposed with an efficient migration. ensures distribution resources VMs that emphasizes timely execution user application by reducing delay. VMP carried out applying Whale Optimization Genetic Algorithm (WOGA), inspired whale evolutionary optimization non-dominated sorting based genetic algorithms. performance evaluation for static dynamic comparison recent state-of-the-arts observed notable reduction in shared servers, cost, consumption time up 28.81%, 25.7%, 35.9% 82.21%, respectively utilization 30.21%.

Language: Английский

---

esDNN: Deep Neural Network Based Multivariate Workload Prediction in Cloud Computing Environments

ACM Transactions on Internet Technology, Journal Year: 2022, Volume and Issue: 22(3), P. 1 - 24

Published: March 14, 2022

Cloud computing has been regarded as a successful paradigm for IT industry by providing benefits both service providers and customers. In spite of the advantages, cloud also suffers from distinct challenges, one them is inefficient resource provisioning dynamic workloads. Accurate workload predictions can support efficient avoid wastage. However, due to high-dimensional high-variable features workloads, it difficult predict workloads effectively accurately. The current dominant work prediction based on regression approaches or recurrent neural networks, which fail capture long-term variance To address challenges overcome limitations existing works, we proposed an e fficient supervised learning-based D eep N eural Network ( esDNN ) approach prediction. First, utilize sliding window convert multivariate data into learning time series that allows deep processing. Then, apply revised Gated Recurrent Unit (GRU) achieve accurate show effectiveness esDNN, conduct comprehensive experiments realistic traces derived Alibaba Google centers. experimental results demonstrate accurately efficiently Compared with state-of-the-art baselines, reduce mean square errors significantly, e.g., 15%. rather than using GRU only. We machines auto-scaling, illustrates number active hosts efficiently, thus costs be optimized.

Language: Английский

---

A Fault Tolerant Elastic Resource Management Framework Toward High Availability of Cloud Services

IEEE Transactions on Network and Service Management, Journal Year: 2022, Volume and Issue: 19(3), P. 3048 - 3061

Published: April 26, 2022

Cloud computing has become inevitable for every digital service which exponentially increased its usage. However, a tremendous surge in cloud resource demand stave off availability resulting into outages, performance degradation, load imbalance, and excessive power-consumption. The existing approaches mainly attempt to address the problem by using multi-cloud running multiple replicas of virtual machine (VM) accounts high operational-cost. This paper proposes Fault Tolerant Elastic Resource Management (FT-ERM) framework that addresses aforementioned from different perspective inducing high-availability servers VMs. Specifically, (1) an online failure predictor is developed anticipate failure-prone VMs based on predicted contention; (2) operational status server monitored with help power analyser, estimator thermal analyser identify any due overloading overheating proactively; (3) are assigned proposed fault-tolerance unit composed decision matrix safe box trigger VM migration handle outage beforehand while maintaining desired level users. evaluated compared against state-of-the-arts executing experiments two real-world datasets. FT-ERM improved services up 34.47% scales down VM-migration power-consumption 88.6% 62.4%, respectively over without approach.

Language: Английский

---

Performance Analysis of Machine Learning Centered Workload Prediction Models for Cloud

IEEE Transactions on Parallel and Distributed Systems, Journal Year: 2023, Volume and Issue: 34(4), P. 1313 - 1330

Published: Jan. 30, 2023

The precise estimation of resource usage is a complex and challenging issue due to the high variability dimensionality heterogeneous service types dynamic workloads. Over last few years, prediction traffic has received ample attention from research community. Many machine learning-based workload forecasting models have been developed by exploiting their computational power learning capabilities. This paper presents first systematic survey cum performance analysis-based comparative study diversified learning-driven cloud models. discussion initiates with significance predictive management followed schematic description, operational design, motivation, challenges concerning these Classification taxonomy different approaches into five distinct categories are presented focusing on theoretical concepts mathematical functioning existing state-of-the-art methods. most prominent belonging class thoroughly surveyed compared. All classified implemented common platform for investigation comparison using three benchmark traces via experimental analysis. essential key indicators evaluated concluded discussing trade-offs notable remarks.

Language: Английский

---

A review on quantum computing and deep learning algorithms and their applications

Soft Computing, Journal Year: 2022, Volume and Issue: 27(18), P. 13217 - 13236

Published: April 7, 2022

Language: Английский

---

Multivariate workload and resource prediction in cloud computing using CNN and GRU by attention mechanism

The Journal of Supercomputing, Journal Year: 2022, Volume and Issue: 79(3), P. 3437 - 3470

Published: Sept. 4, 2022

Language: Английский

---

An AI-Driven VM Threat Prediction Model for Multi-Risks Analysis-Based Cloud Cybersecurity

IEEE Transactions on Systems Man and Cybernetics Systems, Journal Year: 2023, Volume and Issue: 53(11), P. 6815 - 6827

Published: July 10, 2023

Cloud virtualization technology, ingrained with physical resource sharing, prompts cybersecurity threats on users’ virtual machines (VMs) due to the presence of inevitable vulnerabilities offsite servers. Contrary existing works which concentrated reducing sharing and encryption/decryption data before transfer for improving raises computational cost overhead, proposed model operates diversely efficiently serving same purpose. This article proposes a novel multiple risks analysis-based VM threat prediction (MR-TPM) secure minimize adversary breaches by proactively estimating VMs threats. It considers risk factors associated configuration management VMs, along analysis behavior. All these are quantified generation respective score values fed as input into machine learning-based classifier estimate probability each VM. The performance MR-TPM is evaluated using benchmark Google Cluster OpenNebula traces. experimental results demonstrate that computes learns patterns from historical live samples. deployment allocation policies reduces up 88.9%.

Language: Английский

---

A sustainable and secure load management model for green cloud data centres

Scientific Reports, Journal Year: 2023, Volume and Issue: 13(1)

Published: Jan. 10, 2023

The massive upsurge in cloud resource demand and inefficient load management stave off the sustainability of Cloud Data Centres (CDCs) resulting high energy consumption, contention, excessive carbon emission, security threats. In this context, a novel Sustainable Secure Load Management (SaS-LM) Model is proposed to enhance for users with CDCs. model estimates reserves required resources viz., compute, network, storage dynamically adjust subject maximum sustainability. An evolutionary optimization algorithm named Dual-Phase Black Hole Optimization (DPBHO) optimizing multi-layered feed-forward neural network allowing estimate usage detect probable congestion. Further, DPBHO extended Multi-objective secure sustainable VM allocation minimize number active server machines, wastage greener SaS-LM implemented evaluated using benchmark real-world Google Cluster traces. compared state-of-the-arts which reveals its efficacy terms reduced emission consumption up 46.9% 43.9%, respectively improved utilization 16.5%.

Language: Английский

---

Nature-Inspired Intelligent Computing: A Comprehensive Survey

Research, Journal Year: 2024, Volume and Issue: 7

Published: Jan. 1, 2024

Nature, with its numerous surprising rules, serves as a rich source of creativity for the development artificial intelligence, inspiring researchers to create several nature-inspired intelligent computing paradigms based on natural mechanisms. Over past decades, these have revealed effective and flexible solutions practical complex problems. This paper summarizes mechanisms diverse advanced paradigms, which provide valuable lessons building general-purpose machines capable adapting environment autonomously. According mechanisms, we classify into 4 types: evolutionary-based, biological-based, social-cultural-based, science-based. Moreover, this also illustrates interrelationship between well their real-world applications, offering comprehensive algorithmic foundation mitigating unreasonable metaphors. Finally, detailed analysis challenges current promising future research directions are presented.

Language: Английский

---

OFP-TM: an online VM failure prediction and tolerance model towards high availability of cloud computing environments

The Journal of Supercomputing, Journal Year: 2022, Volume and Issue: 78(6), P. 8003 - 8024

Published: Jan. 6, 2022

Language: Английский

---