Analysis of Reduced Quantized Feedback-Based User-Antenna Scheduling Scheme for Multi-user Massive MIMO FDD Systems

Abstract

This article proposes an opportunistic quantized feedback-based user and antenna scheduling (OQFUAS) scheme for multi-user massive multiple-input and multiple-output frequency division duplexing (MU-mMIMO FDD) systems to circumvent feedback bottleneck. As in order to improve system throughput, channel state information (CSI) is mandatory at base stations (BSs). However, with increase in number of antennas as well as users, uplink feedback overhead increases drastically. Hence, user-antenna joint scheduling with limited feedback is essential for MU-mMIMO FDD systems. In this article, we propose the OQFUAS scheme that jointly considers antenna selection and user scheduling (US) utilizing the limited feedback information from users, i.e., antenna indexing and multi-bit signal-to-interference-plus-noise ratio (SINR) quantized information. Particularly in this scheme, the opportunistic users selectively feedback the antenna indexing information, four-bit quantized CSI utilizing the optimal quantization levels and selection thresholds. Furthermore, the closed-form solutions to find the optimal quantization levels and selection thresholds are not tractable (also not discussed in literature) for such networks which depends on a variety of system parameters. This article also presents the mathematical analysis of multi-bit quantization strategy. Moreover, the analytical expressions for achievable system throughput are derived for both heterogeneous and homogeneous MU-mMIMO FDD systems. Interestingly, the analytical results mimic and in accord with simulation results with insignificant gap.