Universal High-Throughput and Low-Complexity LDPC Decoder for Laser Communications

To address the challenges posed by propagation channel impairments and to meet the high data rate requirements of laser communications, this study introduces a pioneering low-density parity-check (LDPC) decoder characterized by its high throughput and low complexity. The unique design of this decode...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE access Ročník 12; s. 33328 - 33336
Hlavní autoři: Kang, Jing, An, Junshe, Zhu, Yan
Médium: Journal Article
Jazyk:angličtina
Vydáno: Piscataway IEEE 2024
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:
Algorithms
Complexity
Computer architecture
Decoders
Decoding
Error correcting codes
Field programmable gate arrays
FPGA
Hardware
high-throughput
Iterative decoding
laser communications
Lasers
LDPC decoder
low-complexity
Message passing
Optical communication
Parity check codes
Pipelining (computers)
Random access memory
Throughput
ISSN:2169-3536, 2169-3536
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:To address the challenges posed by propagation channel impairments and to meet the high data rate requirements of laser communications, this study introduces a pioneering low-density parity-check (LDPC) decoder characterized by its high throughput and low complexity. The unique design of this decoder, based on an inter-frame pipeline and intra-frame parallel (IFPP-IFP) scheme, is specifically tailored to maximize the efficiency of processing units, leading to a substantial increase in decoding throughput. The implementation of IFPP is realized through a novel full-overlap message passing (FOMP) scheme and a dynamic address access (DAA) algorithm, distinguishing it from current solutions. Additionally, the decoder employs a message packing strategy and low-complexity data alignment units to effectively achieve IFP. Compared to existing solutions, our hardware implementation on the Xilinx XCKU060 FPGA demonstrates significant progress. The decoder achieves a decoding throughput of 2.67 Gb/s at 10 iterations and 350MHz. Remarkably, when five decoders are used on a single FPGA device, the throughput soars to 13.3 Gb/s, outperforming state-of-the-art designs by 1.3 times and concurrently reducing resource consumption by half. This combination of resource efficiency and enhanced throughput highlights the innovative and superior nature of our proposed approach.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2024.3371188