Data Recovery with Wide Reader on BPM using ML

Objective: ML-based data detection with wide reader in bit pattern media recording.
Feature: Using machine learning to eliminating 2D inter-symbol interference.

A read head is likely to cover multiple data tracks for reading back signal from a bit patterned media. This project is designed to use a single reader for data recovery even though multiple tracks are read at the same time. A trained machine learning channel should be capable of eliminating both inter-symbol and intertrack interference. Initial findings are very promising.

Machine Learning for Bit Patterned Recording