A load-aware resource allocation and task scheduling for the emerging cloudlet system

Zhang, Feifei; Ge, Jidong; Li, Zhongjin; Li, Chuanyi; Wong, Chifong; Kong, Li; Luo, Bin; Chang, Victor

doi:10.1016/j.future.2018.01.053

Cited by 26 publications

(15 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Otherwise, the task is classified as a latencytolerant. Many research papers classify task types using techniques to identify and classify task characteristics [5][6][7][8]. We define the types of tasks running on a user device, as in Equation (1).…”

Section: Task Analysismentioning

confidence: 99%

“…Although research on MEC has been very active, there are certain limitations in handling real-time services requiring strict adherence to response time standards [4]. In general, an application service can be classified as a latency-sensitive service or a latency-tolerant service according to the urgency level of task end-time violation [5][6][7][8]. A service ought to be distinguished as a latency-sensitive service that must adhere to deadlines or a latency-tolerant service that may be allowed to violate a deadline by an arbitrary amount.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Latency-Classification-Based Deadline-Aware Task Offloading Algorithm in Mobile Edge Computing Environments

Choi

Lee

2019

Applied Sciences

View full text Add to dashboard Cite

In this study, we consider an edge cloud server in which a lightweight server is placed near a user device for the rapid processing and storage of large amounts of data. For the edge cloud server, we propose a latency classification algorithm based on deadlines and urgency levels (i.e., latency-sensitive and latency-tolerant). Furthermore, we design a task offloading algorithm to reduce the execution time of latency-sensitive tasks without violating deadlines. Unlike prior studies on task offloading or scheduling that have applied no deadlines or task-based deadlines, we focus on a comprehensive deadline-aware task scheduling scheme that performs task offloading by considering the real-time properties of latency-sensitive tasks. Specifically, when a task is offloaded to the edge cloud server due to a lack of resources on the user device, services could be provided without delay by offloading latency-tolerant tasks first, which are presumed to perform relatively important functions. When offloading a task, the type of the task, weight of the task, task size, estimated execution time, and offloading time are considered. By distributing and offloading latency-sensitive tasks as much as possible, the performance degradation of the system can be minimized. Based on experimental performance evaluations, we prove that our latency-based task offloading algorithm achieves a significant execution time reduction compared to previous solutions without incurring deadline violations. Unlike existing research, we applied delays with various network types in the MEC (mobile edge computing) environment for verification, and the experimental result was measured not only by the total response time but also by the cause of the task failure rate.

show abstract

Section: Task Analysismentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Latency-Classification-Based Deadline-Aware Task Offloading Algorithm in Mobile Edge Computing Environments

Choi

Lee

2019

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…When the distance is between the RR and CR then the velocity is updated based on Equation (11). Finally, if the velocity is between the range of CR and AR then the velocity changes according to Equation (12)…”

Section: Random Vm Selection Based On Starling Flock For Energy Efmentioning

confidence: 99%

“…Moreover, load balancing and energy consumption is another kind of problem which needs to be addressed while scheduling the tasks to cloud resources [11]. During the schedule of user tasks, load and energy of each resource needs to be realized and decisions can be taken to assign a task to the best VM so that the overall makespan, degree of imbalance and energy consumption is minimized [12].…”

Section: Introductionmentioning

confidence: 99%

Hybrid Starling Social Spider Algorithm for Energy and Load Aware Task Scheduling in Cloud Computing

M¹,

S²

2019

IJITEE

View full text Add to dashboard Cite

The efficiency of the cloud-based systems is greatly relying on the task scheduling algorithm which affects the performance parameters such as makespan, response time, degree of imbalance and cost. In recent years, the energy efficiency is also considered as another challenging issue which affects the efficiency of cloud computing systems. This paper proposes a Hybrid Starling Social Spider Algorithm (Starling-SSA) for Energy and Load Aware Task Scheduling in cloud computing. The Starling-SSA is designed as a hybrid algorithm inspired by the intelligent behavior of social spider and the collective response behavior of starling birds. The foraging behavior of spider is implemented to identify the best VMs for the given task with minimum makespan and degree of imbalance. In addition to this, the distance factor is incorporated inspired by starling flock distance in order identify the closeness of VM pairs and avoids the VMs that are far away, thereby VMs can be limited during the searching process. This will greatly reduce energy consumption by taking only VMs that are belongs to the distance factor. The performance metrics such as makespan, degree of imbalance and energy efficiency are evaluated against the existing algorithms such as EATS, CBAT and HC-ACO. The results presents a significant improvements when comparing to the existing algorithms

show abstract

“…The investigations of the results have provided evidence that this algorithm dominates the existing algorithms when compared with parameters like makespan time and satisfaction rate among the customers in the multi-cloud platform. Feifei et.al [17] proposed an algorithm with a name load-aware resource allocation and task scheduling (LA-RATS). This strategy is used for the cloudlet based MCC system.…”

Section: Related Workmentioning

confidence: 99%

DLMNN: A Deep Learning Modified Neural Network for Balancing the Load of Cloudlets on Cloud

Sambangi,

Gondi

2019

IJEAT

View full text Add to dashboard Cite

Cloud Computing a revolution in the computing world, has enabled the users to utilize the services on the Cloud platform from anywhere at any time. As there is an increase in the demand for the utilization of a cloud environment, there are several challenges to be addressed by the companies or organizations to provide uninterrupted cloud services. To make the cloud services available without interruption, the challenge of balancing the load on cloud servers is a must. Proper allocation of load on the servers optimize the performance of the cloud and improves the efficiency to offer uninterrupted services. Recent studies have shown, cloud always needs to have a capable algorithm to distribute the load on servers of cloud architecture to be available to process cloudlets submitted by the customers. Our paper looks for a new load balancing algorithm that uses the concepts of neural network and is used to allocate the tasks in the cloud. The proposed algorithm consists of two steps. First, Features of tasks and cloud servers are extracted, and the necessary features are selected. The feature selection can be done by using MPCA. In the second step, the selected features are sent as input to the DLMNN algorithm to schedule the task in the cloud. Finally, the experimental results of the proposed DLMNN are compared with some existing algorithms.

show abstract

A load-aware resource allocation and task scheduling for the emerging cloudlet system

Cited by 26 publications

References 26 publications

Latency-Classification-Based Deadline-Aware Task Offloading Algorithm in Mobile Edge Computing Environments

Latency-Classification-Based Deadline-Aware Task Offloading Algorithm in Mobile Edge Computing Environments

Hybrid Starling Social Spider Algorithm for Energy and Load Aware Task Scheduling in Cloud Computing

DLMNN: A Deep Learning Modified Neural Network for Balancing the Load of Cloudlets on Cloud

Contact Info

Product

Resources

About