SYSTEM AND METHOD FOR COMPUTING PARAMETERS FOR A DIGITAL PREDISTORTER

Patent №

US 8,046,199

Granted

2011-10-25

Filed 2006

Owner

TEXAS INSTRUMENTS INCORPORATED

Lab

AI components

1

vision

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

11635903

Digital predistortion system, methods and circuitry for adapting a predistortion system linearizing a non-linear element. The system is a multiply partitioned architecture that addresses long or “memory” effects, and separately addresses shorter duration effects. In a preferred method, the non-linear element is first modeled in software as a nonlinearity and a linearity in cascade form, preferably a Wiener model. The model is validated and adapted to minimize an observed error between the model and the non-linear element. The software model of the non-linear element is then used first to model a predistortion block that addresses short duration effects, and second to separately model a predistortion block that addresses longer duration effects. The models are software executable by an external processor in real time. Periodically the models are executed and used to update the adaptive parameters of the predistortion system without interrupting the system operation.

VisionH03F 1/3247H03F 1/3258H03F 3/24H04L 25/03343H03F 2200/321H03F 2201/3224H03F 2201/3233

AI classification

Vision0.73
Planning0.38
AI hardware0.03
Machine learning0.02
Evolutionary computation0.01
Speech0.00
Natural language0.00
Knowledge representation0.00

Ownership

TEXAS INSTRUMENTS INCORPORATED

assignment · 186910913

Assignors

COPELAND, GREGORY CLARK

On an employer assignment, the assignors are typically the inventors.

© 2026 NYSGPT2525 LLC