ML Data Decoder for Noisy High-Density Optical Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage and retrieval systems face interference issues due to the transduction processes used for encoding and decoding data, which can lead to errors and inefficiencies in reading stored data, especially in high-throughput and high-density storage applications.
Innovation Solution
The implementation of a machine-learning based data decoding method using a convolutional neural network (CNN) that processes component images from an optical storage medium to directly resolve data values without relying on intermediate metrics, optimizing both encoding and decoding processes for improved accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional canonical methods are used to decode data from optical storage medium, then the decoding process follows established procedures, but data retrieval accuracy deteriorates due to interference from transduction processes and noise
Solution Approach 1:
The patent replaces traditional mechanical/optical transduction and decoding processes with a machine-learning-based system. Instead of using canonical optical decoding methods that are susceptible to interference, the invention employs trained neural networks that can learn to ignore or correct for noise and distortion patterns, thereby improving both measurement precision and reliability simultaneously
Solution Approach 2:
The invention changes the fundamental parameters of the decoding process by transitioning from fixed algorithmic procedures to adaptive machine-learning models. The system learns optimal decoding parameters from training data, allowing it to adjust to different noise conditions and storage media characteristics, thus improving accuracy and robustness
2Quantity of substance
If high-density data storage is implemented using optical properties, then storage capacity increases, but interference from transduction processes increases leading to more decoding errors
Solution Approach 1:
The patent replaces traditional optical transduction methods with machine-learning-based decoding. This substitution allows the system to handle high-density storage scenarios where interference is more severe, as the ML models can learn to distinguish actual data signals from transduction-induced noise, thereby maintaining data integrity at high storage densities
Solution Approach 2:
The invention uses training data that copies or simulates real storage conditions including noise and interference patterns. By training on these copied representations of actual operating conditions, the system learns to generalize and accurately decode high-density data even under challenging transduction conditions
Data Source
AI summary
Examples are disclosed that relate to reading stored data. The method comprises obtaining a representation of a measurement performed on a data-storage medium, the representation being based on a previously recorded pattern of data encoded in the data-storage medium in a layout that defines a plurality of data locations. The method further comprises inputting the representation into a data decoder comprising a trained machine-learning function, and obtaining from the data decoder, for each data location of the layout, a plurality of probability values, wherein each probability value is associated with a corresponding data value and represents the probability that the corresponding data value matches the actual data value in the previously recorded pattern of data at a same location in the layout.