User-centric and network-centric radio resource management

Objectives of most radio resource management schemes can be classified as either user-centric or network-centric. User-centric schemes maximize the individual users' interests, while network-centric schemes optimize collective metrics for all users. These two approaches tend to result in qualitatively different resource allocations (with; sometimes, very different degrees of fairness). In this thesis, we consider the joint optimization of both user-centric and network-centric metrics using pricing techniques which mediate conflicts.; First, a joint user-centric and network-centric optimization is considered in a single-cell system. Specifically, a utility function (measured in bits/Joule) and a function of the sum of the throughputs of users in the network are used as the user-centric and the network-centric metrics, respectively. An explicit pricing mechanism is applied to mediate between the user and network objectives. Users adjust their powers in a distributed fashion to maximize their net utilities; while the network adjusts the unit price to maximize its revenue. Characterization of the Nash equilibria and a simple approximation to the optimal solution are presented. Our results show that while users with better channels receive better qualities of service; they also make proportionally higher contributions to the network revenue.; The above joint problem is further extended to a multicell system. In the user-centric optimization, the autonomous base station assignment and power control are formulated as a non-cooperative game among users, while the network prices the resources using two strategies: global pricing that maximizes the revenue and minimax pricing that trades off the revenue for a more even resource allocation.; Finally, we propose a power control problem for a wireless packet network that trades off the individual users' packet-delay requirements against the total (network) energy efficiency. Using the energy per bit cost as a metric of the energy efficiency, defined as the energy consumed per successful information bit transmitted, we minimize the total energy per bit cost in the network subject to an average delay constraint for each user. We consider both centralized and distributed resource allocation strategies that take into account both the buffer dynamics as well as the multiple access interactions. We formulate these strategies as cooperative and noncooperative stochastic games and analyze the resulting operating points using a combination of stochastic game theory and classical queuing theory.