Details
Presenter(s)
Display Name
Danfeng Zhai
- Affiliation
-
AffiliationFudan University
- Country
Abstract
This paper presents a prior-knowledge free modeling method for Nyquist ADCs. Two separate neural networks are proposed to model a different type of distortions in ADC. A new batch generation scheme is proposed to solve the small data set question. The method can cover various static and dynamic distortions at one time without too many settings. The availability of this method can be proved by the above three typical non-ideal ADC designs.