High Throughput Priority-Based Layered QC-LDPC Decoder with Double Update Queues for Mitigating Pipeline Conflicts.

Area of Science:

• Digital Communications
• Error Correction Coding

Background:

• High-throughput decoders are crucial for modern communication systems.
• Quasi-cyclic (QC) low-density parity-check (LDPC) codes are widely used for their error correction capabilities.
• Pipeline stage insertion for increased operating frequency often leads to pipeline conflicts.

Purpose of the Study:

• To develop a high-throughput layered decoder for QC-LDPC codes that mitigates pipeline conflicts.
• To enhance the update frequency of log-likelihood ratios (LLRs) during decoding iterations.
• To improve the overall performance and throughput of LDPC decoders for 5G NR.

Main Methods:

• Implementation of a priority-based layered schedule to maximize LLR updates during pipeline conflicts.
• Introduction of double update queues to further reduce pipeline conflicts and increase LLR update percentage.
• Performance evaluation using 5G NR standards and varying signal-to-noise ratios (SNRs).

Main Results:

• The proposed decoder significantly reduces performance loss compared to state-of-the-art methods (0.6 dB to 0.2 dB).
• Double update queues increase the percentage of updated LLRs per iteration.
• Decoder throughput improved up to 2.85 times at an SNR of 5.9 dB.

Conclusions:

• The developed priority-based layered schedule and double update queues effectively address pipeline conflicts in QC-LDPC decoders.
• The proposed decoder architecture offers substantial throughput gains and performance improvements for 5G NR.
• This work provides a viable solution for high-throughput decoding in advanced communication systems.