UMAS Learning Requirement for Controlling Network Resources

Abstract

This paper presents an intelligent User Manager Agent System (UMAS) which has capability of making management decisions for balancing the network load with the users’ requests for accessing network resources. A conventional users’ request for accessing network resources depends on the rigid rules setting and it can affect the overall performance of network services. UMAS provides additional measure in controlling the network services availability and responsiveness. In providing a different level of services to users, UMAS is required to perform an appropriate learning activity that can furnish an input for a decision of time allocation for a single session. This paper demonstrates the use of Neuro Fuzzy Logic for performing the learning that will be integrated into the UMAS.