Title:Energy-Efficient Nonstationary Spectrum Sharin

Abstract:In this paper, we develop a novel design framework for energy-efficient spectrum sharing in cognitive radio networks, where primary users (PUs) and secondary users (SUs) aim to minimize their transmit power levels subject to their minimum throughput requirements. Most existing works proposed stationary spectrum sharing policies, in which the users transmit at fixed power levels as long as the system parameters (e.g. channel gains) remain unchanged. Since the users transmit simultaneously under stationary spectrum sharing policies, in order to fulfill the minimum throughput requirements, they need to transmit at high power levels due to multi-user interference. To improve the energy efficiency, we construct nonstationary spectrum sharing policies, in which the users can transmit at time-varying power levels even when the system parameters do not change. Specifically, we focus on a class of nonstationary spectrum sharing policies in which the users transmit in a TDMA fashion. Due to the absence of multi-user interference during the transmission, this class of policies can greatly improve the spectrum and energy efficiency compared to existing policies. In addition, the proposed policies have the following desirable properties. First, the policies can be implemented by autonomous users in a decentralized manner. Second, the policies can achieve high energy efficiency even when the users have imperfect and limited (binary) feedback about the interference and noise power level from their receivers. Third, they are deviation-proof, namely the autonomous users will find it in their self-interests to follow the policies. Fourth, the framework can be easily extended to the case where users enter and leave the network. Compared to state-of-the-art policies, the proposed policies can achieve an energy saving of up to 80% when the number of users is large or the multi-user interference is strong.