Resource Allocation Technique for Distributed System using Prioritized Class-Based Algorithm

Abstract

Distributed Systems (DS) are the systems in which the components are distributed at different locations in a network and these components interact with each other by passing messages. In DS each system has a memory and a processor pair. DS is cost-effective, more reliable and faster when compared to the conventional computer systems. But there are some issues in the allocation of resources in such systems as the resources are shared over a number of systems apart from its own individual tasks. The current literature developed various techniques to allocate the resources efficiently by using basic algorithms based on local and global messages and tradeoffs. However, these methods may not be effective as they require special storage and maintenance and constraints the resource availability. In this paper, an efficient resource allocation technique is developed using prioritization and class-based algorithm. The tasks are first given priority and then resources are allocated by fixing a certain percentage of bandwidth for each priority group per execution cycle. The simulation results show that the proposed algorithm is efficient in resource allocation when compared to the conventional techniques.