Improving the quality of business and quality of experience in web services through prioritising and scheduling

Abstract

The growth and proliferation of web services has necessitated new perspectives on metrics for such dynamic and open models of interactions between providers and consumers of services. In addition to traditional Quality of Service (QoS) metrics, there is focus on Quality of Experience (QoE) (which pertains to the user's expectations and perception of service levels) and Quality of Business (QoBiz) (which pertains to considerations such as the service provider's revenue/profit). In this paper, we present a pareto-optimal strategy that aims to satisfy users expected service levels while at the same time improving revenue and throughput of the service providers. Our three-stage model Filter-Predictor-Scheduler (FPS) prioritises and schedules service requests such that both QoE and QoBiz levels are maintained. We present extensive experimental validation of our scheduling algorithm to demonstrate its effectiveness.