Artificial neural network approach for Load Balancing in distributed systems

the pace and accessibility of computation in present days are internet dependent, for calculation arranged all the computing device in a grouped that is known as network. For building of network all computing machines like main frame computer, desktop computer, minicomputer are connected to high speed network, and that networked systems are then referred as parallel system or distributed systems. In centralized systems a single computer acts like a controller that has all the responsibilities namely: jobs creation, job migration, monitoring of overloaded system, monitoring of under loaded systems, migration of jobs etc. However, it is clear that the idea of distributing tasks over several hosts has made the problem of efficient task-resource assignment more difficult. This paper implements load migration from overloaded node to under loaded node with the help of artificial intelligence, further under loaded computing node is also searched[2,9]. There emerges efficiency problem from the fact that it is probable to find idle hosts while there are jobs queuing for execution in other hosts. Consequently it would be better to move from heavily loaded hosts to idle hosts. This and other related efficiency problems are solved via load balancing.