Polar Code List Decoding With Reduced Path Sorting Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current polar code decoding methods using the successive cancellation list (SCL) algorithm are inefficient due to high complexity and slow decoding speed, particularly in parallel decoding processes where numerous paths need to be sorted and evaluated.
Innovation Solution
The proposed method reduces decoding complexity by determining a second decoding path for each first decoding path in the SCL parallel decoding process, where the quantity of second decoding paths is less than 2^n, allowing for efficient path selection and reduced sorting complexity, thereby improving decoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional SCL parallel decoding is used to ensure decoding accuracy, then decoding reliability is improved, but decoding complexity increases and decoding speed decreases
Solution Approach 1:
The patent segments the decoding paths into first decoding paths and second decoding paths. First decoding paths are obtained from the (i-1)th step, and second decoding paths are determined based on these first paths. This segmentation allows the system to process and manage paths in a structured manner, reducing the overall complexity of path management while maintaining decoding accuracy through hierarchical path selection.
Solution Approach 2:
The patent determines second decoding paths in advance based on the first decoding paths before final path selection. By pre-determining which paths to keep and which to prune based on path metric values and sorting, the system performs preliminary filtering that reduces the number of paths needing full evaluation, thereby reducing complexity while preserving reliable paths.
2Reliability
If traditional SCL parallel decoding is used to ensure decoding accuracy, then decoding reliability is improved, but decoding speed decreases
Solution Approach 1:
The patent extracts and removes low-probability paths from the set of all possible decoding paths. By calculating path metric values and sorting paths, the system identifies and eliminates paths with low probabilities of being correct. This extraction of unnecessary paths reduces the computational burden and increases decoding speed while maintaining accuracy by preserving high-probability paths.
Solution Approach 2:
Instead of evaluating all possible paths exhaustively, the patent performs partial evaluation by focusing only on the most promising paths. The system determines a specific quantity of second decoding paths based on first decoding paths and their metric values, performing sufficient (but not excessive) path evaluation to ensure accuracy while avoiding the computational waste of evaluating all possible paths.
3Device complexity
If the quantity of second decoding paths is reduced to less than 2^n, then sorting complexity is reduced, but path selection accuracy may be compromised
Solution Approach 1:
The patent changes the parameter of path quantity from the traditional 2^n to a reduced quantity determined by the first decoding paths. By using path metric values as a parameter to guide the selection of second decoding paths, the system maintains accuracy by preserving paths with high metric values while reducing the total number of paths. This parameter change from exhaustive to selective path keeping resolves the contradiction between complexity and accuracy.
Solution Approach 2:
The patent uses path metric values as feedback to guide the selection of second decoding paths. The metric values provide information about the quality of each path, and this feedback is used to determine which paths to retain. The sorting based on metric values ensures that the reduced set of second decoding paths includes the most promising candidates, maintaining selection accuracy despite the reduced quantity.
Data Source
Figure 1~2
Figure 2A~3
Figure 4~5
AI summary
This application provides a decoding method and apparatus and a device. The method includes: obtaining L1 first decoding paths of an (i-1)th group of to-be-decoded bits, where i is an integer, received data corresponds to P groups of to-be-decoded bits, P is an integer greater than 1, 1<i≤P, and L1 is a positive integer; determining a second decoding path corresponding to each first decoding path, where in the L1 first decoding paths, a quantity of second decoding paths corresponding to at least one first decoding path is different from a quantity of second decoding paths corresponding to another first decoding path, and a quantity of second decoding paths corresponding to each first decoding path is less than 2n, where n is a quantity of information bits included in an ith group of to-be-decoded bits; and determining at least one reserved decoding path of the ith group of to-be-decoded bits in second decoding paths corresponding to the L1 first decoding paths, where the at least one reserved decoding path includes a decoding result of the ith group of to-be-decoded bits. In this way, decoding efficiency is improved.