Neural network-based non-linear A/D conversion
Mohamed Amine Bensenouci, Hammoudi Escid, Mokhtar Attari, Ahmed Bensenouci · 2015
This work describes how feed-forward Artificial Neural Networks (ANNs) can perform Analog-to-Digital (A/D) conversion with a linear and non-linear relationship between the analog input and the digital output in order to eliminate the linearization stage without modifying the analog-to-digital converter's elements and architecture. Adding to that, the speed of this A/D converter will not be reduced due to the unchanged conversion algorithm. Simulation for two types of non-linear input has been performed. The results are discussed and a future work is presented.