AI Digital Predistortion for Multi-Impairment MIMO Beam Steering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing predistortion architectures are inefficient in linearizing communication systems with multiple impairments, particularly in MIMO and mMIMO systems, due to frequent recalculations of DPD model coefficients in response to varying operating conditions and environmental factors, leading to signal quality degradation and increased computational burden.
Innovation Solution
An AI-driven linearization method that self-corrects for impairments without continuous feedback, using a single set of DPD coefficients across a wide range of operating conditions, incorporating azimuth and elevation angles, and reducing computational burden by deploying a single DPD actuator for beam steering directions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional DPD architectures recalculate coefficients frequently in response to varying operating conditions, then signal quality is maintained, but computational burden increases and power consumption rises
Solution Approach 1:
The system performs preliminary characterization of the power amplifier across its entire operating range during a training phase, storing pre-computed DPD coefficients for multiple operating points. During actual operation, the system simply looks up and applies the appropriate pre-computed coefficients based on current operating conditions, avoiding the need for real-time recalculation while maintaining signal quality
Solution Approach 2:
The system dynamically selects from multiple pre-computed DPD coefficient sets based on detected operating conditions (power level, temperature, etc.), transitioning between different predistortion models as operating conditions change, rather than using a single static model or continuously recalculating
2Reliability
If multiple DPD actuators are deployed for different beam steering directions in MIMO systems, then linearization accuracy is improved, but device complexity and power consumption increase
Solution Approach 1:
The system develops a unified DPD model that serves multiple beam steering directions simultaneously. By characterizing the power amplifier's behavior across its entire operating range including different beam directions during training, a single DPD actuator can apply appropriate predistortion for any beam direction by selecting from pre-computed coefficient sets, eliminating the need for separate DPD actuators for each direction
3Use of energy by moving object
If DPD coefficients are adapted continuously based on feedback, then PA efficiency is optimized, but system complexity and computational requirements increase
Solution Approach 1:
The system performs all complex coefficient adaptation and optimization during an offline training phase, where DPD coefficients are computed for various operating conditions. During actual operation, the system simply selects from these pre-optimized coefficients based on current operating conditions, achieving PA efficiency optimization without continuous adaptation complexity
Solution Approach 2:
The system uses sensors to automatically detect operating conditions (power level, temperature, etc.) and autonomously selects the appropriate pre-computed DPD coefficient set without requiring complex real-time feedback loops or continuous adaptation algorithms, simplifying the control architecture while maintaining efficiency
Data Source
AI summary
An artificial intelligence (AI) driven transmission system, having a deployed transmitter including a linearizer and power amplifier wherein the deployed transmitter is deployed in an operational configuration in an operational environment. The system includes a processor configured with an input interface to input digitized linearizer signals, the linearizer signals including information carrying signals, and operating conditions parameter signals, other than the information carrying signal representing metrics affecting transfer characteristics of the deployed transmitter over an entirety of the deployed transmitter operating range. The system further including a digital model of the transmitter, for processing the input digitized linearizer signals and for outputting digital model output signals.