A High-Throughput QC-LDPC Decoder for Near-Earth Application

Turbo-like decoding message passing (TDMP) scheduling, which is suitable for quasi-cyclic low-density parity-check (QC-LDPC) codes, can obtain better decoding performance and shorter convergence time than two-phase message passing (TPMP). Normalized min-sum algorithm (NMSA) is used to reduce complex...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:International Conference on Digital Signal Processing proceedings S. 1 - 4
Hauptverfasser: Li, Shixian, Zhang, Qichen, Chen, Yun, Zeng, Xiaoyang
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.11.2018
Schlagworte:
Clocks
Computer architecture
cyclic-shifted identity matrix
Decoder architecture
Decoding
Hardware
Manganese
normalized min-sum algorithm (NMSA)
Parity check codes
quasi-cyclic low-density parity-check (QC-LDPC)
Throughput
turbo-decoding message-passing (TDMP)
ISSN:2165-3577
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Turbo-like decoding message passing (TDMP) scheduling, which is suitable for quasi-cyclic low-density parity-check (QC-LDPC) codes, can obtain better decoding performance and shorter convergence time than two-phase message passing (TPMP). Normalized min-sum algorithm (NMSA) is used to reduce complexity of soft message processing unit. In this paper, we propose a (8176, 7154) QC-LDPC decoder for Consultative Committee for Space Data Systems (CCSDS) standard. In order to avoid memory access conflicts, two cyclic-shifted identity matrices forming a submatrix are processed individually. The decoder is synthesized by TSMC 65nm CMOS process and can achieve a throughput of 4.1Gb/s at clock frequency of 380MHz.
ISSN:2165-3577
DOI:10.1109/ICDSP.2018.8631641