Host load prediction in cloud computing with Discrete Wavelet Transformation (DWT) and Bidirectional Gated Recurrent Unit (BiGRU) network

Dogani, Javad; Khunjush, Farshad; Seydali, Mehdi

doi:10.1016/j.comcom.2022.11.018

Cited by 36 publications

(19 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the service placement problem, the total energy consumption of fog nodes, denoted as 𝐸 𝑇𝑜𝑡𝑎𝑙 , can be calculated using Equation (7) while providing IoT application services. The energy consumption of fog node 𝑓 𝑗 , 𝐸 𝑓 𝑗 , is defined as the sum of the energy consumed during data transmission, 𝐸 𝑓 𝑗 𝑡𝑟𝑎 , and the energy consumed during processing of the services, 𝐸 𝑓 𝑗 𝑝𝑟𝑜 , which is further defined by Equations ( 8) to (10) which define the energy consumption during data transmission and processing of the service, respectively. These Equations ( 1) to (4) provide a way to quantify the energy consumption of fog nodes and enable the development of energy-efficient service placement strategies in a fog-cloud environment.…”

Section: -2-1 Energy Consumptionmentioning

confidence: 99%

“…As the number of IoT devices grows, so does the volume of data created by these devices, necessitating storage and processing systems that are both efficient and scalable. Cloud computing has evolved as a solution to these problems, enabling the analysis and administration of massive quantities of data created by IoT devices [7][8][9][10]. By utilizing cloud computing, IoT applications may make use of flexible and scalable processing and storage capabilities, as well as a pay-as-you-go pricing model that can help reduce expenses [11].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Two-tier Multi-objective Service Placement in Container-based Fog-Cloud Computing Platforms

Dogani

Yazdanpanah

Zare

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Using cloud computing for Internet of Things (IoT) applications necessitates the transmission of all data to the centralized structure of the cloud, thereby leading to an increase in network traffic and service time. Consequently, cloud computing proves impractical for latency-sensitive IoT applications. Fog computing, acting as an intermediate layer between the cloud and IoT, ensures low latency for such applications. The service placement problem, an NP-hard problem that determines which node should host each service, represents one of the major challenges in the fog computing paradigm. While lightweight containers have emerged as a highly efficient virtualization approach, prior research predominantly employed traditional VM-based architecture for service placement in fog computing. Therefore, this study introduces a multi-objective optimization approach for dynamic service placement in container-based fog computing, accounting for cost, latency, and energy consumption. Specifically, we propose a two-tier framework for resource management in fog computing based on Kubernetes. The multi-objective non-dominated sorting genetic algorithm II (NSGA-II) algorithm balances conflicting performance objectives, such as latency, cost, and energy consumption. Empirical results demonstrate that the proposed method outperforms existing state-of-the-art methods.

show abstract

Section: -2-1 Energy Consumptionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Two-tier Multi-objective Service Placement in Container-based Fog-Cloud Computing Platforms

Dogani

Yazdanpanah

Zare

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…J. Dogani et al 17 employed a Bidirectional Gated‐Recurrent Unit (BiGRU), Discrete Wavelet Transformation (DWT), and an attention mechanism to improve the host load prediction accuracy. This hybrid model was evaluated with cluster data sets from Google and Alibaba and the experimental results showed that the proposed method improve prediction accuracy by 3% to 56% compared to a variety of state‐of‐the‐art methods.…”

Section: Related Workmentioning

confidence: 99%

“…The key factor in auto‐scaling is to choose an appropriate VM on a host that needs to be migrated. Most of the existing works lack this factor as an inappropriate decision may give rise to unwanted downfall in the performance such as high load, more energy consumption, and increased costs 17 . This challenges the researchers to consider such techniques that result in performance up‐gradation in terms of requisite performance attributes.…”

Section: Related Workmentioning

confidence: 99%

Efficient Auto‐scaling for Host Load Prediction through VM migration in Cloud

Verma,

Bala

2023

Concurrency and Computation

View full text Add to dashboard Cite

SummaryThe expeditious deployment of Cloud applications and services on wide‐ranging Cloud Data Centres (CDC) gives rise to the utilization of many resources. Moreover, by the increase in resource utilization, virtualization also greatly impacts achieving desired performance. The major challenges in virtualization are detecting over‐utilized or under‐utilized hosts at the right time and the proper scaling of Virtual Machines (VM) on the accurate host. Auto‐scaling in Cloud Computing allows the service providers to scale up or down the resources automatically and provides on‐demand computing power and storage capacities. Effective utilization and autonomous scaling of resources eventually reduce the load, energy consumption, and operating costs. In this paper, an efficient auto‐scaling approach for predicting host load through VM migration has been proposed. The ensemble method using different time‐series forecasting models has been proposed to forecast the approaching workload on the host. Based on this predicted load, different algorithms have been devised to detect over‐utilized and under‐utilized hosts and VMs can be migrated. The designed approach has been validated by experimentation on a real‐time Google cluster dataset. The proposed technique significantly improves average CPU utilization and reduces over‐utilization and under‐utilization. It also minimizes response time, service level agreement violations, and the slighter number of migrations and scaling overhead.

show abstract

“…Cloud computing is the on-demand provision of computing resources as services over the Internet [2]. Companies no longer must procure, configure, or manage resources.…”

Section: Introductionmentioning

confidence: 99%

Towards a Model for Predicting Traffic Flow for Optimizing Resource Allocation and Usage in a Cloud Computing Systems

Sekwatlakwatla,

Malele

2023

Preprint

View full text Add to dashboard Cite

Short-term traffic flow prediction is one of the most important tools for the organisation to manage traffic and properly allocate system resources. Cloud computing requires concentrated computational resources and predictive tools to monitor traffic and integrate innovations such as artificial intelligence and machine learning. This paper uses compares a stepwise linear regression, linear support vector machine, and Gaussian process regression for tools for predicting traffic flow. The results show that stepwise linear regression performs better than other models in predicting traffic flow in Organisational Cloud Computing Systems. The use of a combination of machine learning models with hourly monitoring and resource allocation will be considered for future studies.

show abstract

Host load prediction in cloud computing with Discrete Wavelet Transformation (DWT) and Bidirectional Gated Recurrent Unit (BiGRU) network

Cited by 36 publications

References 58 publications

A Two-tier Multi-objective Service Placement in Container-based Fog-Cloud Computing Platforms

A Two-tier Multi-objective Service Placement in Container-based Fog-Cloud Computing Platforms

Efficient Auto‐scaling for Host Load Prediction through VM migration in Cloud

Towards a Model for Predicting Traffic Flow for Optimizing Resource Allocation and Usage in a Cloud Computing Systems

Contact Info

Product

Resources

About