Convolutional Coding for Nonvolatile Memory Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current nonvolatile memory systems face challenges in maintaining data integrity due to shifting threshold voltage distributions in multi-level cell storage, leading to increased error rates and misread data, especially as the number of logical states per cell increases.
Innovation Solution
Convolutional coding is employed to encode data before storage, allowing for a larger Hamming distance between allowed sequences, thereby increasing the number of detectable and correctable errors, and using a ½ coding rate to generate two encoded bits from one unencoded bit, which helps in maintaining data accuracy even with high error rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multi-level cell storage is used to increase storage capacity, then the number of logical states per cell increases, but the threshold voltage distributions shift leading to increased error rates and data misread
Solution Approach 1:
The patent applies convolutional coding to encode data before storage in the multi-level cell memory. This preliminary encoding action creates a code structure with minimum Hamming distance between allowed sequences, enabling error detection and correction capabilities that compensate for the threshold voltage shifts and reading errors that occur during subsequent memory operations
Solution Approach 2:
The patent introduces convolutional codes as an intermediary layer between the data and the physical storage medium. This intermediary encoding scheme transforms the raw data into a formatted sequence that includes redundant information, allowing the system to detect and correct errors caused by multi-level cell instability without affecting the underlying storage mechanism
2Reliability
If convolutional coding with ½ coding rate is used to increase error detection capability, then the number of encoded bits doubles, but the storage space requirement increases
Solution Approach 1:
The patent employs convolutional coding with a ½ coding rate, which transforms each input bit into two output bits through a systematic encoding process. This parameter change in the code rate provides enhanced error detection and correction capability by creating a minimum Hamming distance of 3 between allowed sequences, enabling the detection and correction of reading errors in multi-level cell memory
Data Source
AI summary
Data are encoded using convolutional coding prior to storage in a nonvolatile memory array, so that errors that occur when the data are read may be corrected even where there is a large number of such errors. Coding rates of less than one increase the amount of data to be stored but allow correction of large numbers of errors.