Resource allocation and cooperative behavior in fading multiple-access channels under uncertainty

The problems of resource allocation and the characterization of cooperative behavior in a two-user fading multiple-access channel (MAC) in the presence of uncertainty about the channel state information at the transmitters (CSIT) are considered. Game-theoretic techniques are used to characterize cooperative behavior and obtain the system operation point from an individualistic perspective. A two-user bargaining problem formulation, with the utility function being defined as the average rate each user can achieve, is used to obtain a fair and efficient operating point. It is shown that for the conventional bargaining problem formulation, the Nash Bargaining Solution (NBS) is very sensitive to the CSIT, and hence may result in considerable performance degradation. A modified bargaining problem formulation, that does not require an explicit modeling of the actual CSIT error, is proposed to address this issue, and shown to provide significant robustness to the system.