Layered Polar Code Graph Encoding for Lower Block Error Rates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Polar codes, while competitive for error correction in 5G wireless communications, face challenges in achieving optimal performance due to limitations in code length and channel capacity, particularly in practical scenarios where infinite code length is not feasible, leading to suboptimal error correction and decoding efficiency.
Innovation Solution
The implementation of layered polar codes, which involve multiple encoding layers with specific kernel operations and graph parameters to optimize sub-channel selection and encoding processes, enhancing the code exponent and reducing block error rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional polar codes are used with finite code length, then encoding complexity remains low, but error correction performance and code exponent are suboptimal
Solution Approach 1:
The patent divides the polar code encoding process into multiple layers, where each layer applies kernel operations to specific subsets of input bits. This segmentation allows the system to achieve better error correction performance through improved code exponent while maintaining manageable encoding complexity by processing bits in organized groups rather than as a monolithic block.
Solution Approach 2:
The patent introduces a layered dimension to the traditional polar code structure, creating encoding layers that operate in parallel or sequence. This dimensional expansion transforms the single-layer encoding process into a multi-layer architecture, enabling superior finite code length behavior without proportionally increasing overall complexity.
2Reliability
If code length is increased to improve channel capacity, then error correction improves, but practical implementation becomes limited by finite code length constraints
Solution Approach 1:
The patent modifies the encoding parameters by introducing layer-specific graph parameters and kernel operation configurations that optimize performance for finite code lengths. By adjusting these parameters across multiple layers, the system achieves better utilization of available channel capacity without requiring excessively long code lengths.
Solution Approach 2:
The patent performs preliminary organization of input bits into layers and groups before encoding, with each layer pre-configured with specific graph parameters. This preliminary structuring enables more efficient encoding of finite-length codes by preparing the data in an optimal format for the encoding process, thereby improving performance within practical code length constraints.
3Reliability
If multiple encoding layers with graph parameters are implemented, then code exponent and error correction performance improve, but encoding process complexity increases
Solution Approach 1:
The patent segments the encoding process into distinct layers, each handling specific input bits with dedicated graph parameters. This segmentation reduces the complexity of any single encoding operation while achieving superior overall performance through the cumulative effect of multiple specialized layers, thereby improving block error rate without overwhelming complexity.
Solution Approach 2:
The patent employs universal kernel operations that can be applied across multiple layers with different graph parameters. These kernel operations serve multiple functions by operating on different bit subsets in different layers, allowing the system to achieve improved code exponent and error correction performance while reusing the same fundamental encoding building blocks.
Data Source
AI summary
In a layered coding approach, a code configuration parameter of a polar code is determined, and encoding graph parameters are determined based on the determined code configuration parameter. The encoding graph parameters identify inputs for one or more kernel operations in each of multiple encoding layers. Information symbols are encoded by applying the one or more kernel operations to the inputs identified in each encoding layer in accordance with the determined encoding graph parameters.