Pragya Sharma

Pragya Sharma

PhD, School of Electrical and Computer Engineering, Cornell University, Prof. Edwin Kan Group

Online Trainable Near-Field Communication (NFC) Reader

Maxim Integrated, Summer Internship

2017

This project implemented machine learning (ML) based digital data demodulation.

  • Programmed an NFC reader model including analog front-end and digital baseband processing in Python. Implemented new modulation protocols on the output of the existing hardware to predict future challenges.
  • Developed intelligent multi-class classification algorithm using a simple neural network architecture to perform digital data demodulation, which achieved a low test error of 1% for optimal conditions, and 11% with low coupling efficiency and high noise.
  • Implemented a real-time trainable setup for testing the experimental data from the reader, resulting in a low test error of < 2%.