A New Fuzzy Algorithm for Dynamic Load Balancing In Distributed Environment

Load Balancing means to distribute the load among the processors to improve the performance of a parallel and distributed system. This project presents a new fuzzy algorithm for dynamic load balancing in distributed environment. This approach characterizes the uncertainty in a distributed system by using the fuzzy set theory , fuzzy control methods and presents a fuzzy based distributed load balancing algorithm that explicitly reflects the effect of uncertainty in the decision making process. The notion of the linguistic variables is used to model the sender and receiver node. In this we have taken the two input variables, ‘load’ and ‘number of heavy load nodes’ and one output variable ‘status of load balance node’. On the basis of decision rules we have concluded the status of load balance node i.e. whether the node is receiver or sender. The decision rules have implemented on the MATLAB. Then the system model is created on network simulation software. Simulation results show the response time.