I. Introduction
An adaptive network service allows client applications to choose the service attributes required of the underlying system infrastructure. In current information networking environments (e.g., E-banking, tele-shopping), the service attributes depict user-level prescription of the expected quality of service (QoS): such as the latency of remote data access and the system overhead incurred therein. The service provider (SP) implements a distributed algorithm (or, protocol), running on different sites of the network infrastructure, to export the client-requested service with a desired QoS.