RF Connection Testing with Machine Learning to Isolate DUT S11
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing test systems struggle to accurately determine the quality of radio frequency (RF) connections between a test system and a device under test (DUT) due to the embedded S11 parameter values from the DUT, which complicates the measurement of return loss (RL) and requires computationally prohibitive circuit modeling.
Innovation Solution
Utilizing machine learning (ML) models trained on reflected RF signals to distinguish and remove DUT contributions, enabling the determination of RF connection quality by leveraging an RF simulator to generate diverse training datasets, thereby avoiding manual data collection and adapting to various DUT types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional circuit modeling methods are used to determine connection quality, then measurement precision can be achieved, but computational complexity becomes prohibitive
Solution Approach 1:
The patent replaces complex circuit modeling computations with machine learning inference. The ML model, trained on simulated connection data, can quickly determine connection quality metrics (return loss, S11 parameters) without performing computationally intensive circuit simulations, thus achieving measurement precision with reduced computational complexity
Solution Approach 2:
The patent creates a computational model (ML model) that copies the behavior of complex RF connection systems. By training the model on simulated data representing various connection scenarios, the model can replicate connection quality assessments without actually performing the complex physical measurements or circuit simulations, thereby reducing computational burden
2Measurement precision
If DUT contributions are not removed from reflected signals, then measurement process is simpler, but measurement precision deteriorates due to embedded S11 parameter values
Solution Approach 1:
The patent extracts and removes the DUT's S11 parameter contribution from the reflected signal measurements. By separating the DUT's inherent reflection characteristics from the connection quality indicators, the system can accurately measure connection return loss without contamination from the DUT's own electrical characteristics, achieving precise measurement while managing processing complexity through targeted signal decomposition
Solution Approach 2:
The patent segments the reflected signal into distinct components: DUT contribution and connection quality contribution. By analyzing and separating these components, the system can isolate the connection quality metrics from the DUT's embedded parameters, enabling precise measurement of connection characteristics without the complicating influence of the DUT's inherent electrical properties
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
The ML-based approach quickly and accurately determines the RF connection quality, including RL and S11 parameter values, improving the accuracy and reliability of RF testing by isolating DUT contributions and reducing computational complexity.
Implementation Method 1
The tester is configured to transmit radio frequency (RF) signals over the connection setup and to capture reflected signals from the connection setup. The reflected signals are based on the RF signals.
Data Source
AI summary
An example system includes a tester configured to test a device under test (DUT) and a connection setup that is connectable to, and disconnectable from, the DUT. The tester is configured to transmit radio frequency (RF) signals over the connection setup and to capture reflected signals from the connection setup. The reflected signals are based on the RF signals. One or more processing devices are configured to use a trained machine learning model to determine a quality of a connection between the test system and the DUT based on the reflected signals.