Research On Pricing Problems In Wireless Communication Networks

The rapid development of wireless access network technology has brought huge profits for mobile operators in recent years. In the foreseeable future, the environment of wireless communication systems will become more complicated: a geographic area or a hotspot would be covered by several wireless access points with different technologies, such as WiFi, WiMAX, cellular networks(3G,4G), femtocells, etc., and these access points could be operated by different service providers. Since the success or failure of a technology are often closely linked to its business model behind, this thesis is dedicated to establish a general pricing model for wireless communication systems in order to explore the economic factors behind the system. Different from the previous works, we take the interactions of different access points into account which generalizes the former pricing works. Because the operators and users are both rational agents, we use the framework of non-cooperative game theory to model the relationship between operators and operators, between operators and users, and between users and users. Under this general pricing framework, we analyze two special scenarios in order to bring us more intuitive thinking.As the first application of the proposed framework, we consider a N-AP uplink CDMA access network pricing model. Suppose that there is a geographic region cov-ered by N wireless APs which provide access services for a large number of end users. We assume that all the APs share the same spectrum and the users can roam among any of the APs freely and suffer no cost. Since we assume that each user always has data to transmit, the traffic model of the problem is static. The congestion function of each AP is linear if the number of end users is large. We derive the analytical expressions for monopoly and oligopoly market, respectively. In addition, we compare the two markets with respective to social welfare and profits of operators.Another application of the proposed framework is a dynamic traffic wireless net-work pricing model. Similarly, we consider a geographic region covered by N wireless APs but they use different spectrums to communicate so that the problem is tractable. We assume that the arrival of data traffic accords to a Poisson process and the length of each packet accords to exponential distribution. We derive the analytical expressions for monopoly and oligopoly market as well and compare them using numerical methods.