查看文章

In a distributed environment the workload on the network has to be managed in such a way that the total throughput of the system can be maximized. For this to happen some of the jobs have to be migrated from one node to another. When, how and where a job has to be migrated depends upon the load balancing algorithm being used. But it is very difficult to precisely describe the behavior of a complex system as there are many factors which influence it. One way to deal with the uncertainty in the behavior of the system is to use fuzzy logic. Fuzzy logic uses the reasoning of the human mind which is not always in the form of a yes or no. The concept of linguistic variables is used to model the state of the system which is imprecise and uncertain. In this work, we have implemented the fuzzy load balancing algorithm and compared the effect of using different defuzzification methods, reported in the literature.