Data network pricing under quality of service (QoS) guarantee: Single class and multiple classes

Over the last two decades, packet-switching networks, because of their efficiency and reliability, have emerged as the popular choice for data transmission in most enterprises. Because of the queueing effects on these networks, the transmission time of an individual packet is usually unpredictable. Therefore, traditional packet-switching networks typically operate on a best effort basis.;Recently, fast packet-switching networks have gained wide acceptance as a means to provide end-users with quality of service (QoS) requirements for mission-critical applications. It is important for managers of information systems to understand QoS and its organizational consequences because some types of applications are more sensitive to delay and require a certain level of service quality to be maintained. Second, not all data are equally important to organizations, so it may be necessary to prioritize mission-critical business units. Of course, in order to provide the QoS guarantee in these situations, one must design and implement a proper pricing scheme. This dissertation studies the optimal usage-based (such as per-packet) pricing scheme in such networks with the QoS guarantee (characterized in terms of an expected per-packet delay). The first part of the dissertation develops the basic model where the expected per-packet delay parameter is assumed to be fixed and exogenously determined as an industry standard. The second part relaxes that assumption and considers the situation where a provider can choose her own QoS parameter. The third part develops a priority pricing scheme with two classes of QoS guarantee.;A key contribution of this dissertation is an explicit incorporation of a “hard” QoS mechanism (in terms of the expected per-packet delay) in the pricing scheme of network services. Theoretically, this study enriches the traditional competitive literature. It is found that, under certain circumstances, consumers may be worse off in a fully competitive market. Therefore, governments might play an important role in fending off the fierce competition in the networking industry. Practically, the findings from this study can be used as a guide to efficiently manage network resources. They prescribe how a network service provider should charge consumers in the presence of negative network externalities.