Abstract: With the advances in technology, faster growth in the use of computer systems has led to increase in the applications that share the resources there by increasing the amount of load .As the demand for resource sharing applications is grown, load balancing approach has become a necessity factor. In a distributed network, the overall performance of the system depends on the effective division of the workload among all the available set of processing nodes. Load balancing is an approach in which the redistribution of load takes place equally among the set of processing nodes in the system. The main objective is to improve the overall performance there by obtaining the best job response time. In general, classification of load balancing algorithms is done into two major types called static and dynamic. In dynamic, the decentralized approach suffers from the communication overhead because of the exchange of information frequently among the processors. In the existing centralized node based load balancing technique, the workload is assigned randomly to the processing nodes. Once allocated, load balancing conditions is checked at every processing node and the job migration is involved. A centralized node based load scheduling and balancing approach is proposed to overcome the limitations of existing centralized load balancing technique by allocating the random arrival of load among the processing nodes based on FCFS (First Come First Served) scheduling algorithm taking into consideration the current status of every processing node. Here, the load balancing conditions are checked only by the central node to overcome the overhead on the processing nodes. In this approach the transfer of job from one node to another is not done in order to overcome the migration overhead and concentrate much on executing jobs rather than transferring. This job scheduling policy is adaptive that considers the dynamic changes in the system and accordingly schedules and balances the load.
Keywords: Load Balancing, Distributed systems, FCFS, Migration, Communication Overhead, Response Time