Ant Colony and Fuzzy Logic Based Strategy for Load Balancing in Parallel Machines

Abstract

High computing devices are consuming large amount of power as well as energy. Although they are giving good performance for engineering applications but energy consumption in high scale is the crucial point. Along with energy consumption, the performance is also very critical factor in any kind of computing environment these days. There are several algorithms already existing which use scheduling approach based on the energy limitations. The problem of this research is to design and develop an algorithm which can effectively cope up with the energy minimization and faster execution of the tasks to balance the load. Objectives of this thesis is to reduce the energy consumption in cloud computing environment, minimize SLA violations, increase throughput, reduce response time and optimize the load balancing. Reducing energy consumption will bring various benefits like operating cost, reliability cost and environment cost etc. This thesis studied the concept of energy consumption on servers using task scheduling as primary factor. Comparative study of various algorithms like ACO, Fuzzy logic and traditional algorithm for energy consumption along with other parameters are also explained. Approach followed for implementing the solution in this thesis includes fuzzy logic followed by ant colony optimization. Mathematical model of proposed solution for energy consumption and optimized load balancing has also been presented in MATLAB using ACO method.