Uplink Flush Indicator for HARQ Retransmission Control

The transport block is segmented into multiple code blocks, and further organized into code block groups. This segmentation allows the system to identify and retransmit only the specific code block groups that contain errors, rather than retransmitting the entire transport block. The code block group level granularity provides an optimal balance between error correction efficiency and resource utilization.

The patent applies local quality by enabling different handling strategies for different code block groups within the same transport block. Some code block groups may be retransmitted with combining enabled, while others may require flushing and retransmission without combining. This localized approach optimizes resource usage by applying the appropriate error correction strategy to each affected segment rather than uniformly treating the entire transport block.

By segmenting the transport block into code block groups, the patent enables precise identification of which segments require retransmission. The signaling overhead is reduced because the feedback mechanism only needs to indicate which specific code block groups failed, rather than triggering a complete transport block retransmission. This segmentation approach minimizes the information that needs to be signaled while maintaining reliability.

The patent introduces new parameters at the code block group level, including the code block group flush indicator (CBGFI) and code block group transmission information (CBGTI). These parameters enable fine-grained control over retransmission behavior, allowing the system to adjust the retransmission strategy based on the specific error conditions detected in each code block group, thereby reducing unnecessary signaling.

The patent segments the transport block into code block groups, enabling independent error assessment for each segment. This segmentation allows the receiver to selectively combine retransmitted code block groups with prior transmissions only when the prior data is known to be valid, preventing corrupted data from propagating to other segments. Each code block group can be handled independently based on its specific error status.

The patent applies local quality by enabling different combining strategies for different code block groups. For code block groups where the prior transmission was successful, combining is enabled to improve decoding success rate. For code block groups where corruption is detected, combining is disabled and flushing is performed instead. This localized quality control prevents corrupted data propagation while maximizing the benefits of combining where applicable.

The patent introduces segmentation at the code block group level, which provides a practical balance between granularity and complexity. By organizing code blocks into groups rather than handling individual code blocks separately, the patent achieves resource efficiency improvements while limiting the increase in device complexity. The code block group structure provides a manageable level of granularity that simplifies the control logic compared to fine-grained individual code block management.

The patent designs the code block group mechanism to serve multiple functions simultaneously: error identification, retransmission control, combining management, and signaling optimization. This multi-functionality reduces the need for separate complex mechanisms for each function, thereby limiting the overall increase in device complexity while achieving resource efficiency improvements through unified code block group level control.