Concatenated Linear and Non-Linear Precoding for 5G MIMO
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In 5G NR MIMO systems, implementing non-linear precoding is challenging due to high complexity, sensitivity to CSI errors, and the need for advanced demodulation techniques, especially with large antenna arrays and IoT devices requiring low power and long battery life.
Innovation Solution
The method involves concatenating linear and non-linear precoding and generating dual demodulation reference signals (DMRS) for efficient data demodulation, using explicit CSI feedback and a dual DMRS structure for both linear and non-linear precoding, allowing for scaling and phase de-rotation of non-linearly precoded data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If non-linear precoding is implemented in 5G NR MIMO systems, then network throughput and MU MIMO performance are improved, but system complexity increases significantly
Solution Approach 1:
The non-linear precoding process is segmented into two distinct stages: a first-stage linear precoder that performs initial signal processing, and a second-stage non-linear precoder that applies interference pre-subtraction. This segmentation allows the system to achieve non-linear precoding performance while managing complexity through modular architecture, where each stage can be independently optimized and implemented.
2Measurement precision
If non-linear precoding is implemented, then demodulation accuracy is improved, but sensitivity to CSI errors increases
Solution Approach 1:
The system performs preliminary channel state information acquisition and explicit feedback before the actual data transmission. The receiver estimates the channel and provides explicit CSI feedback to the transmitter in advance, allowing the non-linear precoder to be configured with accurate channel knowledge. This preliminary action reduces the impact of CSI errors during the actual demodulation process.
3Measurement precision
If dual DMRS structure is used for concatenated linear and non-linear precoding, then data demodulation performance is improved, but signal processing complexity increases
Solution Approach 1:
The patent introduces dual demodulation reference signals as an intermediary mechanism to bridge the gap between the concatenated linear and non-linear precoding stages. The first DMRS is processed through the linear precoder only, while the second DMRS is processed through both linear and non-linear precoders. This intermediary structure enables the receiver to separately estimate and compensate for the effects of each precoding stage, simplifying the overall demodulation process despite the complex transmitted signal structure.
Data Source
AI summary
Systems, methods, apparatuses, and computer program products for non-linear precoding are provided. One method may include combining, by a network node, of linear and non-linear precoding, for example based on specific channel state information acquisition from at least one user equipment, and generating two types of demodulation reference signals for the concatenated linear and non-linear precoding. The method may include multiplexing, by the network node, the two types of demodulation reference signals with data. A first of the two types of demodulation reference signals may be linearly precoded with one linear precoding matrix from a first-stage linear precoder, and another one of the two types may be linearly precoded with both the first-stage linear precoder and a feedforward filter in a second-stage non-linear precoder from the concatenation of linear and nonlinear precoders.