Feedback Optimization Schemes for Downlink MISO-OFDMA Systems With Heterogeneous Users

Abstract

This paper proposes feedback optimization schemes for multiple-input single-output orthogonal frequency-division multiple access (MISO-OFDMA) systems with heterogeneous users. The proposed scheme adjusts the unit and resolution of the feedback to efficiently exploit the long-term channel statistics, including the path loss, delay spread, and mobility. To develop an analytical optimization framework, we first derive the sum-rate of MISO-OFDMA systems when the channel-state information (CSI) at the transmitter is imperfect because of the quantization error, channel variation, and feedback delay. Based on the derived results, we present effective solutions to the sum-rate maximization problem for distributed and centralized scenarios. When the channel statistics of the users are not available at the base station, each user can improve the sum-rate of the system by minimizing the distortion on CSI feedback under a fixed per-user feedback rate. With the channel statistics of the users, the base station can further control the user set and the feedback rate in addition to the feedback granularity. We evaluate the performance of both distributed and centralized schemes over practical wireless environments, where each user has different long-term channel statistics. Simulation results show that both methods significantly improve the performance of MISO-OFDMA systems.