Polar Code Puncturing Control for Reliable Nonvolatile Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Polar code encoding faces challenges in maintaining reliability due to the inherent characteristics of the generation matrix, leading to loss of information word bits when puncturing is applied, which results in reduced system reliability and inability to normally decode information word bits.
Innovation Solution
The solution involves controlling polar code encoding by overlapping incapable bits and frozen bits, adjusting the locations of frozen bits to overlap with incapable bits, and using a controller with a puncturing bit register, location providing unit, frozen bit inserting unit, and generation matrix calculation unit to optimize the selection of bits to be punctured based on polarization exponents and error probabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If puncturing is applied to polar codes to reduce redundancy, then code rate is improved, but information word bits are lost and system reliability deteriorates
Solution Approach 1:
The patent applies preliminary action by pre-identifying incapable bits through polarization exponent calculation before puncturing occurs. The location providing unit determines which bits cannot be transmitted, and the frozen bit inserting unit pre-places frozen bits at these locations. This preliminary identification and preparation prevents information loss during puncturing, resolving the contradiction between achieving high code rate through puncturing and maintaining system reliability.
2Ease of manufacture
If frozen bit locations are fixed in conventional polar codes, then encoding simplicity is improved, but adaptability to puncturing patterns deteriorates
Solution Approach 1:
The patent applies dynamics by making frozen bit locations variable rather than fixed. The location providing unit dynamically determines frozen bit locations based on the puncturing pattern and polarization exponents. When puncturing patterns change, the frozen bit locations are automatically adjusted to overlap with incapable bits. This dynamic adaptation maintains encoding effectiveness across different puncturing scenarios while preserving the systematic structure of polar codes.
3Device complexity
If incapable bits are not identified and overlapped with frozen bits, then encoding process is simpler, but information loss during puncturing increases
Solution Approach 1:
The patent replaces the mechanical approach of uniform frozen bit placement with a calculation-based system using polarization exponents. The generation matrix calculation unit computes polarization exponents for each bit position, and the location providing unit uses these calculations to identify incapable bits. This substitution of calculation-based identification for fixed-position placement accurately determines which bits cannot be transmitted, preventing information loss while managing complexity through systematic computation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An operating method of a nonvolatile memory controller includes selecting bits of code word of a polar code to be punctured; detecting locations of incapable bits of an input word which have a high error probability owing to puncturing based on locations of the bits to be punctured and on a structure of a generation matrix; refreezing the input word by adjusting the location of frozen bits such that frozen bits and incapable bits of the input word overlap; generating input word bits by inserting frozen bits into the information word bits; generating the code word from the input word bits using based on a generation matrix; generating output bits by puncturing the code word based on locations of the bits to be punctured; and transmitting the output bits to a nonvolatile memory device.