Polar Decoding of Reed-Muller Nodes Using Fast Hadamard Transform
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication systems, particularly in 5G NR, face challenges in efficiently decoding polar-coded data due to the complexity of polar coding schemes and the need for improved decoding methods that can handle intermediate nodes with specific configurations of frozen and information leaf nodes.
Innovation Solution
The implementation of a method that applies a Fast Hadamard Transform (FHT) to the log likelihood ratios (LLRs) associated with Reed-Muller nodes, followed by the selection of paths for simplified successive cancellation list (SSCL) decoding, and the calculation of path metrics to optimize decoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If standard polar decoding methods are used, then decoding can be performed, but computational complexity increases and decoding efficiency decreases
Solution Approach 1:
The patent segments the polar code decoding process by identifying and separately processing Reed-Muller nodes (special nodes with specific frozen bit patterns) from general polar code nodes. This segmentation allows the application of optimized FHT-based decoding for RM nodes while using standard methods for other nodes, thereby reducing overall computational complexity and improving decoding efficiency.
Solution Approach 2:
The patent changes the decoding parameter approach by applying Fast Hadamard Transform (FHT) specifically to Reed-Muller nodes instead of using generic polar decoding methods. This parameter change exploits the mathematical structure of RM codes to reduce the computational complexity from exponential to linear in the block size, significantly improving productivity while managing device complexity.
2Measurement precision
If generic polar decoding is applied to all nodes, then uniform processing is simple, but decoding accuracy decreases for special node configurations
Solution Approach 1:
The patent applies local quality by using different decoding strategies for different node types within the polar code structure. Reed-Muller nodes (which have specific configurations of frozen and information bits) receive specialized FHT-based processing to maximize decoding accuracy, while other nodes use standard polar decoding methods, thus achieving high precision without uniform complexity throughout the system.
3Productivity
If Fast Hadamard Transform is applied to Reed-Muller nodes, then decoding efficiency improves, but implementation complexity increases
Solution Approach 1:
The patent segments the code structure to identify Reed-Muller nodes specifically, allowing FHT to be applied only where beneficial. This selective application reduces implementation complexity compared to applying FHT universally, while still achieving significant decoding speed improvements for the identified special nodes.
Data Source
AI summary
An apparatus may be configured to receive a polar-encoded transmission comprising at least one intermediate node associated with a first configuration of frozen leaf nodes and information leaf nodes. The apparatus may further be configured to apply an FHT to a first set of values associated with a first intermediate node of the at least one intermediate node to generate a second set of values associated with the first intermediate node. The apparatus may also be configured to select, based on the second set of values, one or more paths associated with the first intermediate node for a SSCL decoding. The apparatus may further be configured to calculate a path metric for each of the selected one or more paths associated with the first intermediate node.