Competitive Dynamics of QoS Provision in High-Speed Networks: Quality of Service (QoS) provision in high-speed networks carrying a multitude of traffic with diverse QoS requirements is a difficult problem. Defining who should get what (i.e., fairness), in what sense certain QoS assignments are optimal, and how to facilitate efficient allocation mechanisms are some of the fundamental questions. In previous work, Dr. Park has shown the feasibility of using computational markets based on noncooperative game theory and microeconomics to mediate QoS provision for best effort (ABR) traffic in ATM networks. In a nutshell, traffic streams are tagged by priority labels which are then used by switches employing a form of weighted fair queueing to provide differentiated service. Applications (i.e., users) compete for tagging rights via a bidding process. Under certain conditions, in spite of the selfish behavior of individual applications, the system settles into a globally optimal state. Current research is directed at investigating multi-market systems, i.e., internetworks composed of multiple switches and their interactions.