Polar Code BP Decoding Through Conflict VPE Sign Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Belief Propagation (BP) based decoding for polar codes faces performance limitations compared to Successive Cancellation List (SCL) decoding, particularly in terms of decoding latency and throughput, despite its potential for parallel computing and low latency.
Innovation Solution
The proposed Belief Propagation Conflict Search (BPCS) decoding method involves updating right-to-left messages in the BP decoding process by identifying and modifying conflict verification processing elements, which enhances the decoding performance by adjusting message signs and improving the convergence of the decoding algorithm.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional BP decoding is used, then parallel computing capability is maintained, but decoding performance and error correction capability are limited
Solution Approach 1:
The patent performs preliminary identification of conflict verification processing elements before the main decoding operation. By pre-calculating which processing elements may contain conflicts based on frozen bit positions and message signs, the system prepares correction strategies in advance, improving error correction capability without adding significant runtime complexity
Solution Approach 2:
The patent implements a feedback mechanism where decoding results are checked for conflicts, and when conflicts are detected, the message signs are corrected and decoding is repeated. This iterative feedback process continues until successful decoding or maximum iterations are reached, significantly improving error correction capability while maintaining manageable complexity through targeted corrections
2Reliability
If BP decoding is used, then low latency is achieved, but decoding performance is inferior to SCL decoding
Solution Approach 1:
Instead of implementing full SCL decoding which would incur high latency, the patent applies partial correction actions only where needed. By identifying specific conflict verification processing elements and correcting only their message signs, the system achieves performance improvement closer to SCL without the full computational overhead and latency penalty
Solution Approach 2:
The patent changes the sign parameter of right-to-left messages for conflict verification processing elements based on detected conflicts. This parameter modification resolves decoding conflicts and improves performance while maintaining the fast BP decoding framework, achieving better performance with minimal latency increase
3Reliability
If message signs are updated for conflict verification, then decoding performance improves, but computational complexity increases
Solution Approach 1:
The patent applies local quality by treating different processing elements differently. Conflict verification processing elements identified through specific criteria (frozen bit positions, message sign patterns) receive special sign correction treatment, while other processing elements follow the standard BP decoding process. This localized approach improves performance only where needed, minimizing additional computational complexity
Data Source
AI summary
Various embodiments of the present disclosure provide methods and apparatuses for improved belief propagation (BP) decoding. A method performed by a receiver comprises: performing BP decoding on received information; obtaining, in response to the BP decoding being unsuccessful and based on a first table comprising left-to-right messages associated with nodes of a plurality of processing elements (PEs) for the BP decoding and the received information, a second table comprising right-to-left messages associated with the nodes; searching, based on the first table and the second table, a conflict verification processing element (VPE); updating, for the conflict VPE, both signs of the right-to-left messages associated with its right-upper node and right-lower node to be the same and the sign of the right-to-left messages associated with its left-upper node to be positive; updating the second table; and performing the BP decoding based on the updated second table.