Measurement-based admission control in integrated-services networks

Abstract

To satisfy the quality of service requirements of real-time multimedia applications, networks must employ resource reservation and admission control. In this paper, we describe a new Measurement-Based Admission Control (MBAC) algorithm that uses empirical traffic envelopes of the aggregate traffic flow to allocate network resources. Our framework of traffic envelopes provides a robust and accurate characterization of the aggregate flow, capturing the extent to which flows statistically multiplex as well as the temporal correlation structure of the aggregate flow. In estimating applications' future performance, we introduce the notion of a schedulability confidence level which describes the uncertainty of the measurement-based "prediction", based on which we devise techniques to accurately estimate packet loss probability for a buffered multiplexer servicing heterogeneous and bursty traffic flows.