Exploiting Spatial Channel Covariance for Hybrid Precoding in Massive MIMO Systems


We propose another half and half precoding procedure for huge multi-input multi-yield (MIMO) frameworks utilizing spatial direct covariance networks in the simple precoder outline. Applying a regularized zero-driving precoder for the baseband precoding network, we locate an unconstrained simple precoder that expands motion to-spillage in addition to commotion proportion (SLNR) while disregarding simple stage shifter limitations. In this manner, we build up a procedure to plan a compelled simple precoder that imitates the got unconstrained simple precoder under stage shifter limitations. The fundamental thought is to embrace an extra baseband precoding grid, which we call a remuneration network.

We break down the SLNR misfortune due to the proposed half breed precoding contrasted with completely computerized precoding, and figure out which factors significantly affect this misfortune. In the recreations, we demonstrate that if the channel is spatially associated and the quantity of clients is littler than the quantity of RF chains, the SLNR misfortune ends up insignificant contrasted with completely computerized precoding. The fundamental advantage of our technique originates from the utilization of spatial divert networks so that not exclusively is every client’s coveted flag considered, yet in addition the between client impedance is joined in the simple precoder outline.

BASE PAPER: Exploiting Spatial Channel Covariance for Hybrid

Leave a Reply

Your email address will not be published. Required fields are marked *