Extremal Optimization applied to load balancing in execution of distributed programs

Summary Maximization of availability and minimization of the makespan for transaction scheduling in an on‐demand computing system is an emerging problem. The existing approaches to find the exact solutions for this problem are limited. This paper proposes a task scheduling algorithm using ant colony optimization (MATS_ACO) to solve the mentioned problem. In this method, first, availability of the system is computed, and then, the transactions are scheduled using the foraging behavior of ants to find the optimal solutions. We also modify two known meta‐heuristic algorithms such as genetic algorithm (GA) and extremal optimization (EO) to obtain transaction scheduling algorithms for the purpose of comparison with our proposed algorithm. The compared results show that the proposed algorithm performs better than others.

Section: Experimental Evaluationsmentioning

confidence: 99%

“…• Extremal Optimization: Extremal optimization (EO) 38 is a nature-inspired optimization technique. This technique has moderate computational complexity and small memory requirements.…”

Section: Experimental Evaluationsmentioning

confidence: 99%

Section: Experiments With Varying Mean Time To Failurementioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Maximizing availability for task scheduling in on‐demand computing–based transaction processing system using ant colony optimization

Mahato

Singh

2018

“…However it suffers from high task migration time. De Falco et al proposed external optimization processor load balancing method. It achieves low task migration time.…”

Section: Related Workmentioning

confidence: 99%

Magnum opus of an efficient hospitality technique for load balancing in cloud environment

Sakthivelmurugan

Vimala

Britto

2018

Summary The load balancing approach in the cloud environment is an open issue in the recent days. It helps obtain better resource utilization in cloud environment. It also improves the performance of the system. A good load balancing algorithm in cloud ensures no under loading or over loading in a single host. This paper proposes an algorithm named Hospitality Load Balancing (HLB) algorithm, which aims to balance the load among virtual machines (VM) for increasing the performance and throughput. The proposed algorithm handles preemptive or nonpreemptive tasks, which may be independent. It helps to reduce make span and obtain low task migration. The HLB algorithm has been analyzed and compared with plenty of load balancing technique. An experimental result shows that it is an excellent and effective algorithm when compared with the existing load balancing algorithm. It also illustrates the performance of proposed load balancing scheme with analysis and proves that it attains low makespan and low task migration.

Load balanced transaction scheduling using Honey Bee Optimization considering performability in on‐demand computing system

Mahato

Singh

2017

SummaryThe scheduling of load balanced transactions is an emerging problem in on‐demand computing system. This problem can be studied considering several parameters such as resource availability, performance, and reliability. The existing approaches to find the exact solutions for this problem are limited. This paper presents a Honey Bee Optimization (HBO)–based method to solve the mentioned problem. In this method, first, the load of the system is balanced and then the transactions are scheduled using foraging behavior of honey bees to find the optimal solutions. We also modify four known scheduling algorithms such as Ant Colony Optimization (ACO), Hierarchical Load Balanced Algorithm (HLBA), Dynamic and Decentralized Load Balancing (DLB), and Randomized to obtain transaction scheduling algorithms for the purpose of comparison with our proposed algorithm. The compared results show that the proposed algorithm performs better than the modified existing algorithms.