Exploiting CSI-RS Structure for Efficient Fronthaul Functional Split in 5G O-RAN and Low Complexity Channel Estimation

The limited bandwidth of the Fronthaul (FH) network is one of the main challenges faced by the 5G Open Radio Access Networks (O-RAN). The FH network connects the O-RAN Distributed Unit (O-DU) to one or more O-RAN Radio Units (O-RU). In Frequency Division Duplexing (FDD) Massive Multiple-Input and Multiple-Output (MIMO) systems, the FH load increases in proportion to the number of antennas during the transmission of Channel State Information - Reference Signal (CSI-RS) for downlink channel sounding. However, the O-RAN functional split scales the FH capacity with the number of MIMO layers. This paper proposes an alternative to O-RAN functional split (7-2x) for CSI-RS in Downlink Channel Sounding systems. The proposed split distributes the signal processing functions for CSI-RS transmission into two stages. The first stage (CSI-RS base sequence mapping) is implemented in the O-DU, and the second stage (superimposing orthogonal weights) is implemented in the O-RU. Our results indicate that the proposed split significantly reduces the fronthaul CSI-RS transportation load while enabling much simpler processing at O-DU and O-RU compared to the existing O-RAN split. Additionally, the paper proposes a computationally efficient way to implement Least Squares (LS) channel estimation on the User Equipment (UE) side, again by exploiting the structure of the CSI-RS pilots.