User Provided Networks (UPN) is a recent and hot area of research. A substantial amount of research in this area is focused on incentivizing subscribers of Mobile Network Operators (MNO) to make resources such as bandwidth, battery energy and quota available to other subscribers. In general, proposed incentive mechanisms are based on game theory, in particular they heavily make use of potential game theory and aim to improve the total amount of data downloaded. In this thesis, unlike previous studies, we focus on optimizing the use of existing bandwidth and increasing the data rates of the participants. First of all we focus on providing Internet access or higher data rates to the mobile subscribers who have no Internet access or need higher data rates using the resources of neighboring participants. This problem is essentially the most fundamental problem in UPN. In this study, an energy modeling is proposed and a bargaining model is developed based on Rubinstein bargaining scheme. The effectiveness of the mechanism developed is demonstrated according to the metrics -data rate and price per bit- through
numerical studies. Thus, a solution to the most basic problem of UPN is presented. In the second stage of the thesis, development of an incentive mechanism for UPN
with multiple nodes is studied. At this stage, a distribution function is developed that aims the egalitarian sharing of the existing bandwidth among subscribers and then a game is introduced such that the utility function is based on that egalitarian distribution function. This game shows an uncontrolled network structure in which each subscriber tries to maximize her own data rate, in other words, this is the worst case which can be named the case of anarchy. We named this game the local game because each player is rational only to improve their local utility. Another game has also been proposed in which there is a global potential function that aims taking advantage of all available bandwidth for both network operators and subscribers and distributing that bandwidth fairly among subscribers. These two games are compared with each other in terms of total bandwidth exposed and fairness metrics. In the last stage of our thesis a new class of games is introduced which also include special version of the local game from the previous stage. We name this newly introduced class of games as distribution games. In distribution games, we propose a demand proportional distribution function as well as the egalitarian distribution function. In terms of bandwidth exposed, we prove that distribution games are potential games when both egalitarian and demand proportional distribution functions employed. We analyze the distribution games for the price of stability (PoS) and the price of anarchy (PoA). In the final stage, we simulate the distribution games and its generalized version. As a result of the simulations, we demonstrate the effectiveness of distribution games and its generalized version in terms of total bandwidth exposed and fair sharing of the exposed bandwidth. In addition, by proposing distribution games and examining the similarities and differences of distribution games with other games we contribute to game theory. Although the generalized version of distribution games seems to reach Nash equilibrium in all simulations, there is no proof that the generalized version always reaches Nash equilibrium. Thus, this is a new research problem which requires further investigation. This thesis provides holistic solutions based on game theory from the most basic UPN problem to the most complicated one, and while doing this, it contributes new class of games and open research problems to game theory. |