Opportunistic Power Control for Multi-Carrier Interference Channels

Abstract

We propose a new method for opportunistic power control in multi-carrier interference channels for delay-tolerant data services. In doing so, we utilize a game theoretic framework with novel constraints, where each user tries to maximize its utility in a distributed and opportunistic manner, while satisfying the game's constraints by adapting its transmit power to its channel. In this scheme, users transmit with more power on good sub-channels and do the opposite on bad sub-channels. In this way, in addition to the allocated power on each sub-channel, the total power of all users also depends on channel conditions. Since each user's power level depends on power levels of other users, the game belongs to the \emph{generalized} Nash equilibrium (GNE) problems, which in general, is hard to analyze. We show that the proposed game has a GNE, and derive the sufficient conditions for its uniqueness. Besides, we propose a new pricing scheme for maximizing each user's throughput in an opportunistic manner under its total power constraint; and provide the sufficient conditions for the algorithm's convergence and its GNE's uniqueness. Simulations confirm that our proposed scheme yields a higher throughput for each user and/or has a significantly improved efficiency as compared to other existing opportunistic methods.