Survivor Memory Tree Layout for Polar List Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Flash memory devices face data corruption due to spontaneous jumps in memory cells, leading to reliability issues in storing and retrieving data, which existing error correction codes struggle to address effectively.
Innovation Solution
A method for storing survivor data generated during channel polarization code decoding in a memory module using a list successive cancellation decoding technique, where a binary tree configuration is employed to store selected survivor bits, with the number of edges in each level limited by the list size, and log-likelihood ratio (LLR) is used to select the most probable guesses for storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional error correction codes are used to address data corruption in flash memory, then data reliability is improved, but device complexity and resource requirements increase
Solution Approach 1:
The patent segments the decoding process into multiple stages using a binary tree structure, where each node represents a decoding decision point. This segmentation allows the system to manage complex decoding operations in manageable steps, reducing overall device complexity while maintaining high reliability through systematic error correction at each stage.
Solution Approach 2:
The patent introduces a new dimensional approach by organizing survivor data in a binary tree configuration rather than traditional linear or tabular structures. This dimensional change enables more efficient memory addressing and reduces the hardware resources needed for error correction operations.
2Measurement precision
If list successive cancellation decoding is implemented with traditional memory structures, then decoding performance is improved, but resource requirements and device complexity increase
Solution Approach 1:
The patent merges the survivor memory structure with the decoding process itself, eliminating the need for separate registers and multiplexers. By integrating these functions into a unified binary tree-based memory architecture, the system achieves high decoding performance while significantly reducing resource requirements.
Solution Approach 2:
The patent uses copying mechanisms within the binary tree structure to propagate survivor information efficiently through the decoding stages. This copying approach allows the system to maintain multiple candidate solutions (survivors) without requiring proportional increases in hardware resources.
3Ease of manufacture
If traditional memory architecture is used for storing survivor data, then implementation is straightforward, but decoding efficiency and reliability are reduced
Solution Approach 1:
The patent implements a dynamic memory architecture where the binary tree structure adapts its configuration based on the decoding progress and survivor selection requirements. This dynamic reconfiguration enables the system to optimize memory usage at each decoding stage, significantly improving decoding efficiency compared to static traditional architectures.
Data Source
AI summary
A method of storing survivor data generated while decoding channel polarization codes in a memory module includes setting a list size that corresponds to a number of decoder units used to decode the channel polarization codes, inputting a stream of input bits to the decoder units, and sequentially decoding the input bits. Each input bit is decoded using all previous input bits decoded before the each input bit. The method further includes selecting a plurality of survivor bits from among the decoded input bits, and storing the selected survivor bits in the memory module in a binary tree configuration. The number of edges in each level of the binary tree configuration does not exceed the list size.