Virtual Partitioning for resource sharing by state-dependent priorities: Analysis, approximations, and performance for heterogeneous traffic

We consider Virtual Partitioning, a scheme for sharing a resource among several traffic classes in an efficient, fair, and robust manner. The scheme uses state-dependent trunk reservation to give lower priority to classes which are exceeding their predetermined nominal capacity allocation, thus protecting the behaving classes. Motivated by the integration of multiple services in Broadband/ATM networks: we develop efficient computational algorithms to analyze the important case of heterogeneous traffic classes. Extensive numerical experiments are performed to establish the accuracy of the proposed approximations. We also compare the performance of Virtual Partitioning to that of Complete Sharing and Complete Partitioning. The results demonstrate that the scheme is efficient, fair, and extremely robust with regard to traffic perturbations.