RF Power Amplifier Pre-Distortion for Unstable Multi-Channel Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing pre-distortion systems for radio frequency amplifiers are ineffective in multi-channel wireless transmitters where signal conditions are unstable, leading to suboptimal coefficient training and distortion correction.
Innovation Solution
A method for determining pre-distorter coefficients by selectively updating input samples based on a figure of merit calculated from the frequency and amplitude spectrum, ensuring comprehensive frequency band coverage and optimal amplifier characteristics, even in scenarios with frequency hopping or varying amplitudes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional pre-distortion systems are used in multi-channel wireless transmitters, then the system structure is simple, but the coefficient training becomes suboptimal and distortion correction is ineffective under unstable signal conditions
Solution Approach 1:
The system dynamically adapts to unstable signal conditions by continuously monitoring signal characteristics and adjusting pre-distorter coefficients in real-time. The coefficient determination is made dynamic through selective updating based on figure of merit calculations, allowing the system to respond to changing frequency hopping and amplitude variations rather than using static coefficients
Solution Approach 2:
The invention changes the parameters used for coefficient determination by calculating a figure of merit based on frequency and amplitude spectrum analysis. This parameter change enables selective updating of coefficients only when signal conditions warrant it, improving reliability without requiring complete retraining under all conditions
2Adaptability or versatility
If pre-distorter coefficients are trained using all input samples, then the training data is comprehensive, but the system cannot adapt to unstable signal conditions and frequency hopping
Solution Approach 1:
The system applies local quality by treating different portions of the input signal differently based on their suitability for coefficient training. Instead of uniform treatment of all samples, the invention selectively updates coefficients based on local signal characteristics analyzed through figure of merit calculations, allowing high-precision training when conditions are favorable and adaptation when they are not
Solution Approach 2:
The invention introduces feedback through figure of merit calculation that evaluates the quality of input samples before using them for coefficient training. This feedback mechanism determines whether to update coefficients based on current signal conditions, ensuring both adaptability to unstable conditions and maintenance of training accuracy when possible
3Reliability
If the pre-distorter corrects distortion across the entire frequency band, then the frequency band coverage is comprehensive, but the system performance degrades when certain frequency portions are not occupied
Solution Approach 1:
The system applies partial action by updating pre-distorter coefficients only for the frequency portions that are currently occupied and suitable for training, rather than continuously correcting across the entire frequency band. This selective approach improves reliability for active frequencies while reducing unnecessary processing complexity for inactive portions
Data Source
AI summary
Coefficients of a pre-distorter are determined for distortion correction of a signal amplified by a radio frequency amplifier by sampling the signal at the pre-distorter input to generate a plurality of input samples and sampling the signal at the amplifier output to generate a plurality of output samples, each of the output samples corresponding to one of the input samples, calculating a figure of merit for one or more of the input samples on the basis of the one or more input samples, selectively updating a previously selected set of input samples with the one or more input samples on the basis of the calculated figure of merit, whereby to generate an updated selected set and determining the coefficients of the pre-distorter on the basis of the input samples in the updated selected set and the corresponding output samples.