DMRS-BASED UPLINK DIMENSION REDUCTION IN MASSIVE MIMO RADIO UNIT

Smart overview of the Invention

The patent application describes a method for improving the efficiency of massive MIMO systems in wireless communications. It focuses on a network node's ability to obtain demodulation reference signal (DMRS) symbols that are multiplexed with data symbols within a slot. By calculating dimension reduction weights from the DMRS symbols, the network node can perform dimension reduction to produce channel estimates with reduced dimensions. These estimates are then used to calculate beamforming weights, which are applied to form beamformed data symbols.

Technical Field

The disclosure is relevant to the field of communications, specifically wireless communication methods and devices that support massive MIMO technology. Massive MIMO is a key component in 5G networks, characterized by a large number of antennas at the base-station, which significantly enhances spectrum efficiency and cell capacity through advanced beamforming techniques.

Background

Massive MIMO systems have evolved from long term evolution (LTE) to 5G, where they are deployed on a larger scale. The systems use numerous antennas to serve multiple user-layers, improving spectrum efficiency. However, this also increases the fronthaul capacity requirements, prompting the transition from CPRI to eCPRI interfaces. eCPRI allows for more efficient data transport by moving frequency-domain beamforming functions to the radio unit (RU), reducing the fronthaul capacity needs.

Proposed Solutions

Two solutions are proposed to optimize the uplink functional division in O-RAN WG4 specifications. Both solutions transfer DMRS channel estimation and beamforming weight calculation to the RU. Solution A performs equalization in the distributed unit (DU), while Solution B incorporates equalization within the RU, sending SINR information back to the DU for demodulation. These approaches aim to reduce the number of streams to match the number of layers, enhancing performance while minimizing fronthaul bit rate.

Challenges and Improvements

The proposed solutions increase RU complexity by shifting the DMRS channel estimation and beamforming weight calculation to the RU. To mitigate this, a method is suggested to reduce the spatial dimension of the uplink channel, based on sounding reference signal (SRS) channel estimation in the DU. By performing beamforming on DMRS symbols, the channel dimension is effectively reduced, simplifying channel estimation and weight calculation, thus addressing RU complexity concerns.