MISO Branch Feedback for Predistortion Under Multicollinearity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional MISO systems face inefficiencies due to multicollinearity, which complicates the identification of individual contributions from input signals, leading to poor linearization and increased complexity in radio module architectures, especially in high-efficiency Doherty amplifiers.
Innovation Solution
The introduction of uncorrelated noise patterns in each input branch of the MISO system, combined with noise generators and pre-distorters, allows for the separation of contributions and optimization of pre-distortion, enabling better linearization and reduced hardware complexity by breaking multicollinearity and simplifying feedback networks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional feedback networks are used in MISO systems, then the system can provide basic feedback functionality, but the hardware complexity increases and identification precision deteriorates due to multicollinearity
Solution Approach 1:
The patent introduces dummy ports as intermediary elements that mediate between the feedback network and the signal paths. These dummy ports serve as additional measurement points that break the multicollinearity by providing independent feedback signals, thereby improving identification precision without proportionally increasing hardware complexity
Solution Approach 2:
The patent changes the parameter configuration by adding dummy ports with specific impedance values that differ from conventional feedback networks. This parameter change enables the system to resolve multicollinearity issues by creating a non-singular feedback matrix, improving identification precision while maintaining reasonable hardware complexity
2Reliability
If more feedback networks are added to handle multiple input signals, then the feedback functionality improves, but the hardware complexity and loss of energy increase
Solution Approach 1:
The patent merges multiple feedback functions into a unified feedback network structure that utilizes dummy ports. Instead of implementing separate feedback networks for each input signal, the dummy ports enable a combined feedback approach that maintains reliable feedback functionality while reducing the total number of feedback paths and associated energy consumption
Solution Approach 2:
The dummy ports serve multiple functions simultaneously: they provide additional feedback measurement points, break multicollinearity, and enable efficient signal separation. This multi-functionality improves feedback reliability without requiring proportional increases in hardware components and energy consumption
3Ease of manufacture
If conventional pre-distortion is applied without breaking multicollinearity, then the implementation is simpler, but the linearity and manufacturing precision deteriorate
Solution Approach 1:
The patent applies preliminary pre-distortion through dummy ports before the signals enter the main signal paths. This preliminary action breaks the multicollinearity early in the signal chain, enabling subsequent processing stages to operate with improved linearity without significantly increasing overall implementation complexity
Solution Approach 2:
The patent segments the pre-distortion function by implementing it through separate dummy ports rather than as a single complex pre-distorter. This segmentation allows the system to achieve better linearity through distributed pre-distortion while maintaining implementation simplicity through modular architecture
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
It is provided a method for providing feedback to pre-distorters in branches of a MISO system such that the pre-distortion cancels distortions caused by the signal path and the combiner combining the signals from the branches into which input signals are input. The method comprises generating uncorrelated noises and mixing them with the input signals, evaluating the output of the combiner based on the input signals and the noises in order to determine a respective contribution of each input signal to the output of the combiner, and accordingly determining an appropriate pre-distortion. The signal path may apply a non-linear and/or dynamic function on the signal.