Abstract

Load scheduling defines the process of offering, assigning and balancing the load (tasks/ cloudlets to the virtual machines) in the cloud system effectively. The major intention is to minimize the transfer time and the total cost incurs in the load scheduling of the system. The traditional load scheduling techniques necessitate massive amount of resources and mechanisms, which are dynamic in processing, thereby increases the response time, waiting time and the total computation cost. This paper presents an efficient load scheduling technique called hybridization of binary tree optimization with Gravitational Search Algorithm (BTO-GSA) algorithm to minimize the computation time. The total computation time cost comprises of execution cost and transferring cost. It operates on hybrid Splitting Point Selection technique based GSA to search the optimal positions of the particles in the search space. The use of BTO algorithm depends upon the mathematical tree subject and enhances the outcome and searching speed by continuously eliminates the portions of the search space with minimum fitness for minimizing and purifying the search space. The BTO-GSA model has been implemented using CloudSim simulator and a detailed comparative result takes place under several aspects. The simulation outcome indicated that the BTO-GSA algorithm has offered superior performance over the compared methods in a significant way.

Description