Scheduling jobs in computational grid using hybrid ACS and GA approach

Abstract

Metaheuristics algorithms show very good performance in solving various job scheduling problems in computational grid systems. However, due to the complexity and heterogeneous nature of resources in grid computing, stand-alone algorithm is not capable to find a good quality solution in reasonable time. This study proposes a hybrid algorithm, specifically ant colony system and genetic algorithm to solve the job scheduling problem. The high level hybridization algorithm will keep the identity of each algorithm in performing the scheduling task. The study focuses on static grid computing environment and the metrics for optimization are the makespan and flowtime. Experiment results show that the proposed algorithm outperforms other stand-alone algorithms such as ant system, genetic algorithms, and ant colony system for makespan. However, for flowtime, ant system and genetic algorithm perform better.

Minimizing data access latency in data grids by neighborhood‐based data replication and job scheduling

“…The evaluation results indicate that the combination of meta-heuristic algorithms renders better results than the simple metaheuristic algorithms. 26 In Younis and Yang, 27 static schedulers are used for non-dependent jobs. This tactic improves make-span by combining the Variable Neighbor Search (VNS) algorithm with the Genetic and Ant-Colony algorithms.…”

A Loosely Coupled Hybrid Meta-Heuristic Algorithm for the Static Independent Task Scheduling Problem in Grid Computing

“…A hybrid method, called ACO+TS, which combines ACO and TS in a loosely coupled fashion, was proposed in [17]. The experimental results show that the use of TS with ACO improves the quality of the solutions; however, the hybrid method took over 3.5 hours to achieve these results, while the authors in [6] suggested a loosely coupled hybrid meta-heuristic, called ACO+GA, for the task scheduling on computation grid, which combines ACO and GA. However, a non-standard dataset was considered and the code used for their implementation is not available to allow for an even-handed comparison.…”

“…Several stand alone meta-heuristics, such as Ant Colony Optimization (ACO), Particle Swarm Optimization (PSO), Genetic Algorithm (GA), Variable Neighborhood Search (VNS) and Tabu Search (TS), have been applied successfully for various types of scheduling problems. However, the results achieved by these methods could be further improved by combining two or more meta-heuristics [6]. The resulting new high-level algorithm would then inherit the best features of the combined meta-heuristics.…”

Hybrid meta-heuristic algorithms for independent job scheduling in grid computing

“…Meta-heuristic approaches such as the Genetic Algorithm (GA), Tabu Search (TS), Particle Swarm Optimisation (PSO), Variable Neighbourhood Search (VNS) and Ant Colony Optimisation (ACO) have all proven their effectiveness in solving different scheduling problems. However, hybridising two or more meta-heuristics shows better performance than applying a stand-alone approach [4]. The new high level meta-heuristic will inherit the best features of the hybridised algorithms, increasing the chances of skipping away from local minima, and hence enhancing the overall performance [23].…”