Polar Code Belief Propagation With Prioritized Kernel Updating
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Polar codes exhibit poor performance for small code sizes, such as N = 512, compared to other coding techniques like Turbo codes or LDPC codes, and require optimization for Binary Discrete-input Memory-less Channels (BDMC), with high computational complexity in decoding due to the need for extensive beliefs propagation.
Innovation Solution
A method for improving beliefs propagation in polar code decoding by computing a performance-improvement metric for each kernel, sorting kernels based on this metric, and iteratively updating beliefs to prioritize kernels with the most relevant information, thereby optimizing the decoding process while meeting a stop condition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If beliefs propagation is performed on all kernels in polar code decoding, then decoding accuracy is improved, but computational complexity increases significantly
Solution Approach 1:
The patent segments the set of all kernels into two groups: stable kernels (where input and output beliefs are in agreement) and unstable kernels (where beliefs differ). By dividing the decoding process into these segments, the method selectively applies full beliefs propagation only to unstable kernels while skipping stable ones, thereby reducing computational complexity while maintaining decoding accuracy for the critical unstable portions.
Solution Approach 2:
The patent applies partial action by performing beliefs propagation on only a subset of kernels (those that are unstable) rather than all kernels. This partial application of the decoding operation achieves sufficient decoding accuracy by focusing computational resources on the kernels that most need updating, avoiding the excessive computation required to process already-stable kernels.
2Ease of operation
If kernels are processed in arbitrary order during beliefs propagation, then implementation is simpler, but decoding performance deteriorates
Solution Approach 1:
The patent performs preliminary action by computing a performance-improvement metric for each kernel before the beliefs propagation process. This metric identifies which kernels are unstable and would benefit most from processing. By pre-classifying kernels based on this metric, the method establishes an optimal processing order that prioritizes unstable kernels, thereby improving decoding performance while maintaining clear implementation steps.
Solution Approach 2:
The patent introduces dynamics by making the kernel processing order adaptive rather than fixed. The processing sequence is dynamically determined based on the computed performance-improvement metrics and stability status of each kernel. This dynamic ordering allows the decoder to adaptively focus computational effort on the most critical kernels during each iteration, improving performance without requiring complex reordering logic during execution.
Data Source
Figure 1~2B
Figure 3~4B
Figure 4C~4D
AI summary
For performing beliefs propagation in a scope of polar code decoding, the polar code having being based on a structure of L sub-polarization stages of N/2 parallel kernels Ki,j , the kernels at each sub-polarization stage being separated from their neighbour kernels by a shuffler, a decoder performs : computing (S501) a value (i,j) of a performance-improvement metric for each kernel Ki,j; and sorting (S502) the kernels in a list in decreasing order of the values (i,j) The decoder then performs a beliefs propagation iterative process as follows: updating (S503) output beliefs for the W top kernels of the list , and propagating said output beliefs as input beliefs of the neighbour kernels of said W top kernels; updating (S504) output beliefs for each neighbour kernel of said W top kernels following update of their input beliefs, and re-computing (S505) the performance-improvement metric value (i,j) for each said neighbour kernel; setting (S505) the performance-improvement metric for said W top kernels to a null value; and re-ordering (S506) the kernels in the list . Then, the decoder repeats the beliefs propagation iterative process until a stop condition is met.