Polar Code Bit Distribution for Lower-Latency ML Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Polar codes suffer from poor minimum distance at lower code sizes, leading to increased decoding latency in wireless communication systems, particularly in next-generation wireless systems like 5G, due to the complexity and latency associated with maximum likelihood (ML) search spaces in decoding trees.
Innovation Solution
The proposed solution involves strategically distributing redundant outer-code bits within the polar decoding tree to reduce the search space size of internal nodes, allowing for more efficient ML decoding by maximizing the degree and quantity of ML nodes, thereby reducing decoding latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If maximum likelihood decoding is applied to polar codes, then decoding accuracy is improved, but decoding latency increases due to large search space
Solution Approach 1:
The patent segments the polar decoding tree into multiple levels and identifies specific internal nodes for ML decoding. Instead of applying ML decoding uniformly across the entire tree, the method selectively applies it to particular internal nodes based on their contribution to decoding accuracy, thereby reducing the overall search space while maintaining essential decoding performance.
Solution Approach 2:
The patent applies different decoding strategies to different parts of the decoding tree. ML decoding is applied to selected internal nodes where it provides the most benefit, while other nodes use simpler decoding methods. This local differentiation optimizes the balance between accuracy and latency by concentrating computational resources where they are most needed.
2Loss of time
If the search space of ML decoding is reduced, then decoding latency is decreased, but decoding accuracy deteriorates
Solution Approach 1:
The patent performs preliminary analysis to identify which internal nodes in the decoding tree contribute most to decoding accuracy. By pre-determining the optimal set of internal nodes for ML decoding based on code size and coding rate, the system prepares a streamlined decoding path that maintains accuracy while reducing latency, avoiding the need for exhaustive search at all nodes.
3Reliability
If code size N is increased, then polar codes achieve better performance approaching Shannon capacity, but minimum distance deteriorates at lower values of N
Solution Approach 1:
The patent dynamically adjusts decoding parameters based on the code size N and coding rate. For smaller code sizes where minimum distance is poor, the method selectively applies ML decoding to critical internal nodes to compensate for the inherent weakness. This parameter adaptation allows the system to maintain reliable performance across different code sizes without being constrained by the poor minimum distance property of small polar codes.
Data Source
AI summary
Certain aspects of the present disclosure generally relate to techniques for encoding, generally including obtaining a payload, determining a set of internal nodes to distribute one or more non-payload bits to based, at least in part, on a target maximum likelihood (ML) search space size for internal nodes in a polar decoding tree, a search space size of each of the internal nodes, and an available number of the non-payload bits left to distribute, forming an information stream by interleaving the non-payload bits with bits of the payload by, for each internal node in the set of internal nodes, assigning one or more non-payload bits to one or more leaf nodes in a subtree rooted at that internal node in the set of internal nodes, and generating a codeword by encoding the information stream using a Polar code.