An Intelligent Agent for Negotiating Qos in Priced Abr Connections

We investigate the usage of an Intelligent Agent at the user-side in an environment where users are served by and charged for ABR connections. We formulate an optimisation problem pertaining to the case where bandwidth allocation is proportionally fair, and is based on the amount of money users are willing to pay per unit time. Each user is assumed to select his willingness-to-pay so as to maximise his utility function; the optimal selection varies with the demand (i.e. total willingness-to-pay) by other users and with the congestion in the network. The objective of the Intelligent Agent at the user-side is to continuously tune the amount of money ooered to the network, by learning the user behavior; thus, this amount will be adapted to the variations of demand and congestion. In this paper, we analyse the main issues related to the above problem, and we propose a simple, yet promising algorithm for the Intelligent Agent.