scholarly journals An Optimization on Task Scheduling for Makespan, Energy Consumption, and Load Balancing in Cloud Computing Using Meta-Heuristic

Author(s):  
Fajar Kusumaningayu
2022 ◽  
Vol 12 (1) ◽  
pp. 0-0

Fog computing and Edge computing are few of the latest technologies which are offered as solution to challenges faced in Cloud Computing. Instead of offloading of all the tasks to centralized cloud servers, some of the tasks can be scheduled at intermediate Fog servers or Edge devices. Though this solves most of the problems faced in cloud but also encounter other traditional problems due to resource-related constraints like load balancing, scheduling, etc. In order to address task scheduling and load balancing in Cloud-fog-edge collaboration among servers, we have proposed an improved version of min-min algorithm for workflow scheduling which considers cost, makespan, energy and load balancing in heterogeneous environment. This algorithm is implemented and tested in different offloading scenarios- Cloud only, Fog only, Cloud-fog and Cloud-Fog-Edge collaboration. This approach performed better and the result gives minimum makespan, less energy consumption along with load balancing and marginally less cost when compared to min-min and ELBMM algorithms


2019 ◽  
Vol 8 (4) ◽  
pp. 3040-3049

Cloud computing is widely used resource sharing computational technology to provide fast, reliable, and scalable computational process for organizations and companies without the need to build and maintain their own server. The research area about cloud computing is dynamic and versatile. One may have concern on the privacy, security, networking, optimization, etc. Due to huge demand for cloud computing, it creates several problems such as makespan, energy consumption, and load balancing. Task scheduling is one of the technologies that have been applied to solve those objectivities. However, task scheduling is one of the well-known NP-hard problems, and it is difficult to find the optimum solution. In order to solve this problem, previous studies have utilized meta-heuristic method to find the best solution based on the solution spaces. This study proposed Particle Swarm Optimization (PSO) to solve the multi-objective task scheduling to achieve the optimum solution. The effectiveness of the proposed algorithm will be compared with Genetic Algorithm (GA), Clonal Selection Algorithm (CSA), and Bat Algorithm (BA). This study converts three objectivities into single objectivity optimization with each objectivity act as variable assigned with weight that present its priority and has implemented those meta-heuristics. The simulation result from ten data set shows that PSO able to outperform GA, CSA, and BA especially for makespan and energy consumption without the cost of algorithm duration since PSO has fast convergence rate compare to the other three algorithms and making it a good choice for dynamic task scheduling in data center cloud computing where the algorithm duration is one of important factor


2019 ◽  
Vol 8 (4) ◽  
pp. 10093-10099

Recently, the rapid development in processing speeds, fast storage devices and better network connectivity, hasaccelerated the popularization of cloud computing. Cloud computing is an on-demand-servicewhich provides users with high end servers,storage and processing capabilities where the user need not be concerned with its infrastructure.Although, there are abundant resources in the cloud infrastructure, for the efficient working and execution of tasks, task scheduling plays a crucial role. Task scheduling results in better performance (throughput) of the system along with better resource utilization which ultimately results inreduced energy consumption. At any given time, a processor should never be in idle state, as it still consumes some amount of energy. In this paper, the use of Quantum Genetic Algorithm has led to the reduction in energy consumption. The objective is to find a scheduling sequencewhich can be implemented ina cloud computing environment. Along with minimizing energy consumption, the algorithm helps reduce makespan time of a processor as well.The results show a decrease in energy consumption by 10-15% under different test scenarios involving a variable number of tasks, processors, and the number of iterations (generations) for which the algorithm was run. The algorithm converges to the desired result within 10-15 iterations, as can be seen from the results published in this paper.


Booking figuring is reliably a fervently issue in appropriated processing condition. Remembering the true objective to take out system bottleneck and modify stack logically. A stack changing endeavor booking count in light of weighted self-assertive and input frameworks was proposed in this paperFrom the outset the picked cloud masterminding host picked assets by necessities and made static estimation, and some time later coordinated them; other than the tally picked assets from which composed by weight self-confidently; by then it got standing out powerful data from effect burden to channel and sort the left. Finally it accomplished oneself adaptively to structure stack through information systems. The examination demonstrates that the calculation has stayed away from the framework bottleneck adequately and has accomplished adjusted burden and furthermore self-flexibility to it.keywords: Task Scheduling; Feedback Mechanism; Cloud Computing; Load Balancing


2020 ◽  
pp. 1042-1057
Author(s):  
Xiaojing Hou ◽  
Guozeng Zhao

With the wide application of the cloud computing, the contradiction between high energy cost and low efficiency becomes increasingly prominent. In this article, to solve the problem of energy consumption, a resource scheduling and load balancing fusion algorithm with deep learning strategy is presented. Compared with the corresponding evolutionary algorithms, the proposed algorithm can enhance the diversity of the population, avoid the prematurity to some extent, and have a faster convergence speed. The experimental results show that the proposed algorithm has the most optimal ability of reducing energy consumption of data centers.


Author(s):  
Xiaojing Hou ◽  
Guozeng Zhao

With the wide application of the cloud computing, the contradiction between high energy cost and low efficiency becomes increasingly prominent. In this article, to solve the problem of energy consumption, a resource scheduling and load balancing fusion algorithm with deep learning strategy is presented. Compared with the corresponding evolutionary algorithms, the proposed algorithm can enhance the diversity of the population, avoid the prematurity to some extent, and have a faster convergence speed. The experimental results show that the proposed algorithm has the most optimal ability of reducing energy consumption of data centers.


Sign in / Sign up

Export Citation Format

Share Document