CSI Feedback Compression with Limited Downlink Pilots for 5G FDD-NR Massive MIMO Systems

In this article, the Channel State Information (CSI) feedback in Frequency Division Duplexing (FDD) with imposed constraints on the number of CSI Reference Signals (RS) is analyzed. In this regard, Massive MIMO systems have a clear necessity for a scalable optimization of the whole CSI feedback process as the number of transmission ports increases and a large amount of radio resources is required for pilot mapping. Our goal is to study the possibility to utilize a lower number of CSI-RS pilots by applying a proper beam-based port mapping to reduce the number of CSI ports while evaluating the impact on the CSI related procedures and resulting performance. Specifically, we analyze how CSI algorithms in the context of future feedback mechanisms for 5G New Radio (NR) could be implemented under these constraints and how to get a good channel reconstruction with tolerable losses. In simulation results, we show the performance of a CSI compression scheme using a limited number of CSI-RS pilots in comparison with the current 3GPP Rel. 15 and Rel. 16 Type II standards with full CSI-RS pilots availability.