LDPC Basegraph Channel Coding for Low-Latency Wireless Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication systems face challenges in achieving high performance and efficient channel coding, particularly in supporting increased data throughput and reliability requirements for emerging applications like enhanced mobile broadband, massive machine-type communication, and ultra-reliable low-latency communication, where existing coding methods struggle with latency and error correction.
Innovation Solution
The implementation of a method and device for channel coding based on a low-density parity check (LDPC) matrix, utilizing specific basegraph structures and encoding techniques to enhance decoding efficiency and error flow management in wireless communication systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If LDPC basegraph with dual diagonal submatrices is used for channel coding, then decoding efficiency is improved, but device complexity increases
Solution Approach 1:
The LDPC basegraph is segmented into multiple submatrices (T_BG, D_BG, A_BG, B_BG, C_BG, E_BG) with specific structural patterns. The dual diagonal structure of T_BG and D_BG submatrices enables efficient decoding by dividing the large sparse matrix into manageable blocks that can be processed in parallel, improving decoding efficiency while controlling complexity through structured segmentation.
Solution Approach 2:
The patent applies parameter changes by using circular permutation matrices with specific shift values (e.g., Z=16, 32, 64) to construct the basegraph submatrices. By varying the permutation parameters and matrix dimensions while maintaining the dual diagonal structure, the system achieves different code rates and block lengths, optimizing decoding performance for different communication scenarios without fundamentally changing the decoding architecture.
2Productivity
If LDPC encoding is implemented to support high data throughput, then communication capacity increases, but latency increases
Solution Approach 1:
The patent uses pre-defined basegraph structures with dual diagonal submatrices that are designed to enable efficient encoding and decoding operations. The systematic arrangement of submatrices (information part A_BG, C_BG and parity part B_BG, D_BG, T_BG, E_BG) allows preliminary organization of parity-check relationships, reducing the computational burden during actual encoding and decoding operations, thereby lowering latency while maintaining high throughput capability.
Solution Approach 2:
The LDPC code is segmented into systematic parts (information bits in A_BG and C_BG submatrices) and parity parts (B_BG, D_BG, T_BG, E_BG submatrices). This segmentation allows separate processing of information and parity generation, enabling parallel computation paths that increase data throughput while reducing overall encoding latency through pipelining.
3Reliability
If error correction capability is enhanced for reliable communication, then reliability improves, but decoding complexity increases
Solution Approach 1:
The patent implements local quality by creating different submatrix structures with specific properties within the overall basegraph. The dual diagonal structure of T_BG and D_BG submatrices provides localized error correction capability for specific bit positions, while the systematic arrangement of A_BG, B_BG, C_BG, and E_BG submatrices optimizes overall error correction performance. This localized structural design enables efficient belief propagation decoding by reducing the number of non-zero elements that need to be processed, thereby improving reliability without proportionally increasing decoding complexity.
Data Source
AI summary
A method for transmitting an information block on the basis of a low density parity check (LDPC) code in a wireless communication system, according to the present disclosure, may comprise: encoding an information block on the basis of a LDPC basegraph H_BG including [MATRIX]; and transmitting the encoded information block. Each element of H_BG is either zero (“0”) or one (“1”), and each element which is “0”, among the elements of H_BG, may represent a Z×Z zero matrix, and each element which is “1”, among the elements of H_BG, may represent a Z×Z matrix acquired on the basis of a circular permutation matrix acquired by circularly shifting a Z×Z identity matrix to the left or right. The submatrix T_BG of H_BG may be a dual diagonal matrix, and the submatrix D_BG of H_BG may be a dual diagonal matrix. The encoding of the information block on the basis of H_BG may comprise encoding the information block on the basis of a parity check matrix (PCM) H which corresponds to H_BG.