Design and Simulation of Adaptive Measurement-based Admission Control Algorithms for Controlled-load Service

The purpose of call admission control in Controlled-load Service network is to offer a Quality of Service (QoS) commitment to guarantee that QoS bounds are not violated. The traditional way of achieving this goal by declaring the worse case traffic descriptors for the incoming calls results in poor bandwidth utilization. In recent years, measurement-based admission control has become an appealing alternative to improve bandwidth utilization. It exploits the statistical gain and offers adaptivity to changing traffic condition. The measurement-based admission control first proposed by Jamin et. al in [1] was then improved by Casetti et. al in [2] with an adaptive measurement window in the algorithm. Building on the work of [2], this paper examines the algorithm design and simulation of adaptive sampling time in the algorithm. The proposed algorithms are tested through simulations under difference traffic scenarios and proven to produce higher level of utilization without violating the delay-based QoS guarantees. The simulation results are compared with the algorithm described in [2]. Key-Words: Bandwidth Optimization, Admission Control, QoS-enabled Internet