Duplexer Impedance Tuning Using ML Leakage Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Determining the impedance of an antenna in radio frequency devices is challenging, leading to inefficient isolation between transmitter and receiver signals, as existing impedance tuners often overshoot or mismatch, resulting in poor isolation performance.
Innovation Solution
The implementation of a radio frequency device with a voltage standing wave ratio (VSWR) detector and machine-learning models to accurately determine antenna impedance, adjusting impedance tuner settings based on measured leakage values and reflection coefficients, and using multiple smaller machine-learning models for efficient impedance matching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional impedance tuning methods are used, then the impedance tuner can be adjusted, but it often overshoots or mismatches the antenna impedance, resulting in poor isolation performance
Solution Approach 1:
The patent implements a feedback mechanism where the system measures transmitter leakage and receiver leakage, uses these measurements to estimate antenna impedance through a machine learning model, and adjusts the impedance tuner accordingly. This closed-loop feedback system continuously optimizes the impedance matching, preventing overshoot and mismatch issues inherent in traditional open-loop tuning methods.
Solution Approach 2:
The patent replaces traditional mechanical or manual impedance tuning mechanisms with a machine learning-based estimation system. Instead of relying on conventional tuning procedures that may overshoot, the system uses trained machine learning models to predict optimal impedance settings based on leakage measurements, achieving more precise and reliable impedance matching.
2Measurement precision
If multiple test impedance settings are applied to determine antenna impedance, then measurement accuracy improves, but the tuning process becomes more complex and time-consuming
Solution Approach 1:
The patent employs pre-trained machine learning models that have already learned the relationship between leakage patterns and antenna impedance characteristics during the training phase. During actual operation, the system only needs to perform a single or minimal number of leakage measurements and apply the pre-trained model, eliminating the need for multiple iterative test settings and significantly reducing tuning time while maintaining high measurement accuracy.
3Measurement precision
If a single large machine-learning model is used for impedance matching, then accuracy is maintained, but computational complexity and processing requirements increase
Solution Approach 1:
The patent divides the impedance matching task into multiple independent stages, each handled by a specialized machine learning model. One model estimates antenna impedance from leakage measurements, while another determines optimal impedance tuner settings based on the estimated impedance. This segmentation reduces the complexity of individual models, making them more computationally efficient and easier to implement while maintaining overall system accuracy.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables efficient and accurate determination of antenna impedance, improving isolation between transmitter and receiver signals by precisely matching impedance tuner settings, reducing unnecessary tuning adjustments and enhancing communication quality.
Implementation Method 1
a voltage standing wave ratio detector coupled to the duplexer and the antenna. The voltage standing wave ratio detector determines an antenna impedance at the antenna.
Data Source
AI summary
This disclosure provides techniques for impedance matching. A radio frequency (RF) device includes a power detector to determine a transmitter leakage and a post-processing unit to determine a receiver leakage, and determines if isolation is acceptable based on the leakages. The RF device may include a device for measuring antenna impedance. Otherwise, the RF device may select multiple tuner settings (e.g., capacitor values) for test signals to be transmitted and received at a target frequency, determine multiple sets of leakage values, determine multiple reflection coefficients based on the multiple sets of leakage values, and determine an estimated antenna impedance at the target frequency based on the reflection coefficients. The RF device then determines impedance tuner settings based on the measured or estimated antenna impedance. Alternatively, the RF device determines impedance tuner settings using an inverse machine-learning model based on a determined matching impedance.