Performance Analysis of Particle Swarm Optimization Algorithms for Jobs Scheduling in Data Warehouse

S. Krishnaveni, M. Hemalatha


The Enterprise Information System contains data warehouses frequently residing in the number of machines in at least one data center. Many jobs run to bring data into the data warehouses. Jobs are scheduled by dependency basis. Data processing jobs in the data warehouse system involve many resources, so it is necessary to find the best job-scheduling methodology. Particle Swarm Optimization (PSO) is used to find solutions easily and fruitfully, so it is employed in several optimization and search problems. Improved PSO, Hybrid Improved PSO (Improved with simulated annealing (SA)) and Hybrid PSO (three neighborhood SA algorithms are designed and combined with PSO) are used to achieve better solutions than PSO. This paper shows the use of hybrid improved PSO to scheduling multiprocessor tasks, Hybrid PSO to minimize the makespan of job-shop scheduling problem for each best solution that particle find. Algorithms are demonstrated by applying in benchmark job-shop scheduling problems. The superior results indicate the successful incorporation of PSO and SA. This survey results shows the optimal scheduling algorithm to reduce runtime and optimize usage of resources.

Keywords: HPSO, Hybrid ImPSO, ImPSO, Job Scheduling, Particle Swarm Algorithm, Simulated Annealing

Full Text:




  • There are currently no refbacks.

Copyright (c) 2016 International Journal of Advanced Research in Computer Science