A Fuzzy Rule Based Load Balancing Model for a Distributed Service Process Engine

Abstract

A distributed engine with a fuzzy rule based load balancing model for service process is presented. A heuristic algorithm is employed to evaluate and then indicate the potential load of a client's request with a load tag firstly. Based on the load tag of an incoming client's request and load statuses of the engine, the controller of the service process engine allocates the request to an appropriate service according to a set of fuzzy rules. Algorithms and it's structure are discussed in detail; in addition, experiment results are provided to demonstrate the feasibility of the proposed approach.

Dynamic Handling for Cooperating Scientific Web Services

“…Some recent works in the area of distributed computing and fuzzy logic include load balancing for a distributed service process engine [5]. In [6] fuzzy logic is used in master/slave application approach on MPI clusters to balance the work units among slaves.…”