Performance Evaluation of Routing Protocols in VANETs

Abstract: Grid computing is an emerging technology in distributed systems that provide a high performance computing platform to solve larger scale applications by coordinating and sharing computational power, data storage and network resources. All Grid resources are heterogeneous. A grid integrates and coordinates resources and users of different administrative domains inside the same company or in different countries. Task scheduling is one of the key research areas in grid computing. Scheduling is the NP Complete problem which allocates resources for the incoming tasks to the resource broker. Scheduling problem may become tedious when the resources are dynamic and the size of Grid in increased. The goal of scheduling is to achieve highest possible system throughput and to match the application’s need with the available computing resources. This paper primarily focuses on scheduling the tasks to the available resources in the Grid environment. The resource allocation process is based on Artificial Bee Colony (ABC) algorithm. The objective of this algorithm is to generate optimal solution dynamically. By this scheduling, the task is completed in minimum time and uses the available resources in efficient manner. In proposed method Single Shift Neighborhood (SSN), Double Shift Neighborhood (DSN), Ejection Chain Neighborhood (ECN) and Genetic Algorithm (GA) techniques are applied for migration of jobs to the resources and the result are compared based on makespan and resource utilization. The result shows that Genetic Algorithm technique will give better results when compared to Single Shift and Double Shift Neighborhoods. This heuristic algorithm provides an optimal task scheduling in all heterogeneous computing environments.

Keywords: Grid Computing, Task Scheduling, Heuristic Algorithms, Makespan, ABC, SSN, DSN, ECN, GA.