Investigating The Use Of Game Theoretic Distributed Framework Design For Interference Communication Systems

Resource allocation problem in a multi-user MIMO Multiple-input Multiple-outputcommunication system using noncooperative games are mostly based on centralized frame-works where a central resource manager performs all the resource allocation functions. Theinefficiency of centralized frameworks especially in modern wireless networks employingthe MIMO technology posed the need for a non-centralized approached called the Distribu-tive frameworks in which resource allocation decisions are performed by all the nodes in thenetwork. Interactive decision making process obscures the decision making resulting in con-flicts among the various decision makers. Game theory has been proposed as an incentiveto solve the difficulty imposed by optimization method.This research is in two folds.The first part is based on a no-ccoperative scenario whichpredominantly deals with Multiuser MIMO. We overview the waterfilling results existingin the literature from a unified viewpoint and propose an algorithm that bridges the gapbetween a wide family of waterfilling solutions and their efficient implementation in prac-tice; to be more precise, we provide a practical algorithm to evaluate numerically a generalwaterfilling solution, which includes the currently existing waterfilling solutions and othersthat may possibly appear in future problems. We consider the channel link information ratesand the maximization of these information rate using transmit power as a constraint in adistributive approach using game theory. The Nash Equilibrium (NE) provides the solutionof the existence and uniqueness of the game. We then provide a mathematical frameworkproviding sufficient conditions guaranteeing the uniqueness of the NE and the convergenceof waterfilling based on the analysis of the proposed partially asynchronous distributed al-gorithms with channel state information (CSI). The proposed scheme is an extension a.k.arobust scheme described in[79]and the proposed mathematical framework is also instrumen- tal to study the extension of the game to the more general MIMO case.The second part of this thesis addresses the cooperative scenario that focuses on the co-operativeness of the users in a relay transmission environment. We design a game-theoreticapproach for cooperative relay networks using Amplify-and-Forward as a relaying proto-col under spectral mask constraints (SMCs). The cooperative strategy employed as game-theoretic approach is based on the Nash bargaining solution (NBS). The Nash equilibriumand Nash bargaining solutions are derived for a two-player game based on the joint timedivision multiplexing and frequency division multiplexing (TDM/FDM) as a mode of coop-eration. The condition for the uniqueness and existence of the NBS is derived.As main contributions, we establish the existence, uniqueness and convergence of theNE for the proposed games.