Distributed MMSE-IRC Processing for Fronthaul Bottleneck
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The capacity of the fronthaul link in massive MIMO aided cloud radio access networks is bottlenecked by the increasing data traffic as the number of antennas and digital transceiver units at remote radio units (RRUs) increases, leading to performance degradation due to inaccurate interference and noise covariance estimation in existing MMSE-IRC processing techniques.
Innovation Solution
The implementation of distributed minimum mean-square error interference rejection combining (MMSE-IRC) processing, where the RRU performs a first phase of receive beamforming and compresses signals using a modified extended channel, and the BBU performs a second phase of regularized zero forcing, utilizing a linear combination of sample and structured covariance matrices to reduce interference and noise covariance estimation errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the number of antennas and digital transceiver units at RRU is increased to achieve massive MIMO gains, then the network capacity is improved, but the data traffic that needs to be transferred through FH link increases proportionally, causing FH link capacity bottleneck
Solution Approach 1:
The MMSE-IRC processing is segmented into two phases: first phase (receive beamforming and interference covariance estimation) executed at RRU, and second phase (signal detection) executed at BBU. This segmentation allows local processing at RRU to reduce the volume of data that must be transmitted over the fronthaul link, thereby resolving the capacity bottleneck while maintaining network performance
Solution Approach 2:
The patent transforms the processing architecture from a centralized single-dimension model to a distributed two-phase model operating across different spatial and functional dimensions. By distributing processing tasks between RRU and BBU, the system optimizes fronthaul utilization without compromising overall network capacity
2Device complexity
If conventional MMSE-IRC processing is used at BBU, then processing is simplified, but interference and noise covariance estimation becomes inaccurate due to limited feedback information from RRU
Solution Approach 1:
The RRU performs preliminary interference covariance matrix estimation using locally available received signals before transmitting data to the BBU. This preliminary action provides the BBU with more accurate interference statistics, enabling better interference rejection in the second phase of processing without significantly increasing overall system complexity
Solution Approach 2:
The interference covariance matrix estimated at the RRU acts as an intermediary information carrier, bridging the gap between limited fronthaul feedback and the BBU's need for accurate interference statistics. This intermediary enables accurate interference suppression at the BBU without requiring extensive direct measurements
3Extent of automation
If all MMSE-IRC processing is performed at BBU, then centralized control is maintained, but fronthaul link capacity is exceeded due to the need to transmit all raw received signals
Solution Approach 1:
The processing pipeline is segmented into distributed phases, with the RRU handling the computationally intensive first phase of receive beamforming and interference estimation locally, thereby reducing fronthaul data volume. The BBU retains centralized control for the second phase of signal detection, maintaining automation while respecting fronthaul capacity constraints
Data Source
AI summary
Various embodiments herein provide techniques for minimum mean-square error interference rejection combining (MMSE-IRC) processing of a received signal, distributed between a baseband unit (BBU) and a remote radio unit (RRU). The RRU may perform a first phase of processing based on an extended channel that includes a channel of one or more user equipments (UEs) served by the RRU and interference samples that correspond to other cells or additive noise. The first phase may include scaling the interference samples by a scaling coefficient to obtain a modified extended channel, and performing maximum ratio combining (MRC) on the modified extended channel to obtain a processed signal. The RRU may send the processed signal to the BBU for the second phase of processing. The second phase of processing may include regularized zero forcing to remove interference. Other embodiments may be described and claimed.