CBG Header for Fast RLC PDU Deliveries in 5G NR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In 5G NR networks, the conventional TB-based ARQ mechanisms are inefficient due to their low granularity, leading to delays and increased buffer sizes, as correctly received MAC/RLC PDUs are delayed while waiting for failed CBs or CBGs to be retransmitted and decoded, especially since MAC/RLC PDUs are not aligned with CB or CBG boundaries.
Innovation Solution
The introduction of a CBG header in each code block group (CBG) that indicates the starting location of the first protocol data unit (PDU) within the CBG, allowing for immediate forwarding of decoded PDUs even if previous CBGs have been missed, reducing the need for large buffers and minimizing delays.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If TB-based ARQ mechanism is used, then transmission reliability is improved, but transmission delay increases and buffer size increases
Solution Approach 1:
The transport block is segmented into multiple code block groups (CBGs), allowing selective retransmission of only the failed CBGs rather than the entire TB. This segmentation enables finer-grained error handling, reducing the time loss when only part of the TB fails to decode correctly.
Solution Approach 2:
Different CBGs within the same TB can be handled with different quality of service. Successfully decoded CBGs can be forwarded immediately with high priority, while failed CBGs are marked for retransmission. This local quality differentiation resolves the contradiction by allowing reliable transmission of good data while managing retransmission of bad data separately.
2Reliability
If TB-based ARQ mechanism is used, then transmission reliability is improved, but buffer size increases
Solution Approach 1:
By segmenting the TB into CBGs, the buffer only needs to store failed CBGs for retransmission rather than the entire TB. This reduces the quantity of data that must be buffered, as successfully decoded CBGs can be forwarded immediately without waiting for potential retransmissions of other CBGs.
Solution Approach 2:
The buffer management is optimized by treating each CBG independently. Only the failed CBGs require buffering for retransmission, while successful CBGs are processed and forwarded immediately. This local quality approach minimizes the total buffer size required while maintaining transmission reliability.
3Productivity
If CBG-based ARQ is used, then ARQ efficiency is improved, but MAC/RLC PDU delivery is delayed when CBs are not aligned with CB boundaries
Solution Approach 1:
The patent applies local quality by differentiating between successful and failed CBGs at the CBG level, while maintaining PDU-level forwarding capability. When a CBG is successfully decoded, the corresponding PDU can be forwarded immediately even if other CBGs in the same TB fail. This resolves the contradiction by allowing efficient ARQ handling at the CBG level while enabling timely PDU delivery based on local success.
Solution Approach 2:
The patent segments the TB into CBGs and further identifies PDU boundaries within CBGs. This dual segmentation allows the system to retransmit only failed CBGs while forwarding successfully decoded PDUs, thereby improving ARQ efficiency without delaying PDU delivery when alignment issues occur.
Data Source
AI summary
A method in a sending node of a communications network includes: encoding a transport block, TB, including data of at least one protocol data unit, PDU, to generate a code block group, CBG, comprising one or more code blocks; defining a CBG header indicative of a starting location of a first PDU within the CBG; and transmitting the CBG including the CBG header. A method in a receiving node includes: receiving one or more code block groups, CBGs, each CBG comprising a CBG header indicative of a start location of a first protocol data unit, PDU, within the CBG; attempting to decode each received CBG; responsive to failing to decode a first CBG, attempting to decode a second CBG, and responsive to successfully decoding the second CBG: identifying the start location of the first PDU in the second CBG; buffering data of the second CBG prior to the identified start location; and forwarding data of PDUs following the identified start location.