LDPC Code Block Segmentation for Dual Kmax Selection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The existing code block segmentation methods for LDPC codes in 5G NR are inadequate for handling two maximum information block lengths, leading to performance degradation and increased latency, particularly in URLLC scenarios, as they cannot effectively reuse the code block segmentation design from LTE.

Innovation Solution

A method where a base station or terminal determines whether to use the maximum length of the first or second predetermined information bit for code block segmentation based on signaling indication or code rate identification, segmenting the transport block accordingly, with the maximum length of the first predetermined information bit being greater than the second, to optimize code block segmentation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If a single code block segmentation method is used for both large and small base graphs, then the segmentation design can be simple and reusable from LTE, but the decoding performance degrades and latency increases for short information blocks

Engineering Contradiction:
Improvecode block segmentation design simplicityVSAvoiddecoding performance
Core Design Contradiction:
Ease of manufactureVSReliability

Solution Approach 1:

The patent segments the code block segmentation method into two distinct approaches: one for large base graphs (Kmax=8448) and another for small base graphs (Kmax=2560). This allows each segmentation method to be optimized for its specific base graph type, preventing performance degradation while maintaining design clarity through separation of concerns

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces dynamic selection of code block segmentation strategy based on the determined base graph type. The system dynamically adapts the segmentation approach (using different maximum information block lengths) according to whether a large or small base graph is selected, enabling optimal performance for each scenario without sacrificing overall system simplicity

Inventive Principle:
Principle #15Dynamics

2Ease of operation

If a single code block segmentation method is used for both large and small base graphs, then the implementation can be straightforward, but the latency increases particularly for short information blocks in URLLC scenarios

Engineering Contradiction:
Improveimplementation straightforwardnessVSAvoiddecoding latency
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

The patent divides the segmentation implementation into two clear paths corresponding to large and small base graphs. Each path has its own optimized maximum information block length (8448 or 2560), allowing short information blocks to use the smaller base graph with appropriate segmentation, thereby reducing latency while keeping each implementation path straightforward

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the maximum information block length parameter based on the base graph type determination. When a small base graph is selected for short information blocks, the maximum length parameter is adjusted to 2560, enabling faster processing and reduced latency compared to using the fixed LTE segmentation approach

Inventive Principle:
Principle #35Parameter changes

3Quantity of substance

If the maximum length of the first predetermined information bit is used for code block segmentation, then the system can handle larger transport blocks, but the code rate optimization is compromised for scenarios requiring lower code rates

Engineering Contradiction:
Improvetransport block sizeVSAvoidcode rate optimization
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent dynamically selects between two maximum information block length parameters (8448 and 2560) based on the base graph type determination. This dynamic parameter selection allows the system to use the larger maximum length (8448) when handling large transport blocks with large base graphs, while switching to the smaller maximum length (2560) when using small base graphs that require optimized lower code rates, thus optimizing both scenarios

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP3654557B1Code block segmentation method, terminal, and base station
Publication Date: 2022.04.20 DATANG MOBILE COMM EQUIP CO LTD
  • EP3654557B1 patent drawingFigure 1~2
  • EP3654557B1 patent drawingFigure 3~4
  • EP3654557B1 patent drawingFigure 5~6

AI summary

Provided are a code block segmentation method, a terminal, a base station, and a computer-readable storage medium. The code block segmentation method includes: a base station determining whether to use the maximum length of a first pre-set information bit for code block segmentation or to use the maximum length of a second pre-set information bit for code block segmentation; if it is determined to use the maximum length of the first pre-set information bit for code block segmentation, the base station segmenting a transport block into one or more segments by taking the maximum length of the first pre-set information bit as an upper limit; and if it is determined to use the maximum length of the second pre-set information bit for code block segmentation, the base station segmenting a transport block into one or more segments by taking the maximum length of the second pre-set information bit as an upper limit, wherein the maximum length of the first pre-set information bit is greater than the maximum length of the second pre-set information bit. According to the embodiments of the present disclosure, the technical problem of code block segmentation when there are two maximum information block lengths can be effectively solved.