Reduced Redundancy Indexing for Beam-Time Delay CSI Compression in 5G FDD Massive MIMO Systems

The study of Channel State Information (CSI) schemes for FDD Massive MIMO is a hot topic in recent years due to the imperative need to tackle the constraints in terms of CSI uplink feedback overhead. An alternative for having a CSI reduced representation is to exploit sparsity, for example, by extracting the respective correlations in space and frequency domain, which makes it possible to transmit a selected number of taps from a beam-time delay channel representation. In this article, our goal is to study the possibility of exploiting the fact that locations of such taps change relatively slow for low speed User Equipments (UE). In that aspect, one of the issues related is the bit overhead from the indexing mechanism which is needed in order to locate such taps on the CSI beam-time delay grid for channel reconstruction in the Base Station (BS). Our main contribution in this article is to provide a method for reduced redundancy indexing which provides adaptive CSI report payload by reducing unnecessary overhead from indexing location of repeated taps from previous CSI reports. In order to verify its robustness, we study in our analysis the functioning of this technique in a certain range of velocities and different CSI compression configurations.