Power Amplifier Memory Modeling for Multi-Waveform Linearization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing power amplifier models are limited in their ability to compensate for non-linearity and memory effects across varying input waveforms and power levels, leading to inadequate performance in amplifying signals with different characteristics.
Innovation Solution
A universal memory-based model is extracted using a specific extraction waveform with tones of varying frequencies, ensuring the model can be used for a range of waveforms and power levels without the need for additional signal extraction, and digital pre-distortion is applied to compensate for non-linearity and memory effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a power amplifier model is extracted using a specific input waveform and power level, then the model can accurately compensate for non-linearity and memory effects at that specific condition, but the model cannot adequately compensate for input signals having substantially different waveform and power level
Solution Approach 1:
The patent applies universality by designing an extraction waveform that can serve multiple functions: it characterizes the power amplifier's non-linear behavior across a wide range of operating conditions. The waveform includes multiple frequency components and power levels, enabling a single model extraction process to capture amplifier behavior under diverse conditions, making the model universally applicable rather than condition-specific
Solution Approach 2:
The patent utilizes parameter changes by varying multiple parameters of the extraction waveform simultaneously - frequency, power level, and temporal characteristics - to comprehensively characterize the power amplifier's behavior. By changing these parameters during model extraction, the system captures the amplifier's non-linear and memory effects across different operating points, enabling the model to adapt to various input conditions without requiring separate extractions
2Measurement precision
If multiple models are extracted for different waveforms and power levels, then each model can accurately compensate for its specific condition, but the hardware complexity and power consumption increase
Solution Approach 1:
The patent eliminates the need for multiple separate models by creating a single universal model that handles all waveform and power level conditions. The extraction waveform is designed to probe the amplifier's behavior across the entire operating range, allowing one model to replace what would otherwise require multiple condition-specific models, thereby reducing hardware complexity
Solution Approach 2:
The patent merges multiple model extraction requirements into a single unified extraction process. By combining multiple frequency components, power levels, and waveform characteristics into one comprehensive extraction waveform, the system obtains all necessary characterization data in a single measurement campaign, avoiding the need for separate extraction processes and associated hardware for each condition
3Measurement precision
If multiple models are extracted for different power levels, then each model can accurately compensate at its specific power level, but the power consumption and time required for model extraction increase
Solution Approach 1:
The patent implements continuous useful action by designing an extraction waveform that continuously probes the power amplifier across multiple power levels within a single uninterrupted measurement sequence. Rather than performing discrete, separate extractions at each power level, the system captures the amplifier's behavior across the entire power range in one continuous process, significantly reducing the total time required
Solution Approach 2:
The patent applies preliminary action by incorporating all necessary power level variations into the extraction waveform design before the measurement begins. The waveform is pre-configured to include the full range of power levels and conditions needed for comprehensive model extraction, eliminating the need for time-consuming iterative adjustments and separate measurement sessions for each power level
Data Source
AI summary
A method for modelling a power amplifier, including memory effect modelling, for general input waveforms and power levels involves generating an extraction waveform having a plurality of tones each having a different frequency, a difference between the frequencies of two adjacent tones of the plurality of tones not being an integer multiple of a difference in frequency between any two other adjacent tones of the plurality of tones. The method further involves providing the extraction waveform to the power amplifier, receiving output from the power amplifier generated in response to the extraction waveform, and generating a model of the power amplifier based on the output.