Invention:
This invention discloses a rate-adaptive forward error correction (FEC) scheme based on spatially-coupled (SC) LDPC codes derived from quasi-cyclic (QC) LDPC codes with its field programmable gate array (FPGA) architecture. FPGA emulation demonstrates that the proposed LDPC codes provide larger coding gain and smaller error floor compared to the QC-LDPC base code with comparable computational complexity. The hardware friendly structure of the codes make them a promising candidate for the next-generation intelligent optical communication systems such as long-haul optical transmission system.
Background:
Low-density parity check (LDPC) codes were discussed as early as 1962, in information theory research. The codes have encountered various performance barriers to speed and accuracy when decoding high volume of information, such as Shannon limit performance when information is decoded using a probabilistic algorithm, but continuing developments in new code construction, such as binary spatially-coupled codes based on Euclidean geometry have reduced errors compared with regular LDPC codes. As the speed requirement of modern optical communication systems keeps increasing, forward error correction (FEC) becomes an essential technique to enable high-speed transmission ranging from long-haul to access networks. LDPC codes have long strived for FEC from LDPC block code to the most recent spatially coupled (SC) LDPC code. SC-LDPC code approached asymptotic capacity as LDPC convolutional code (LDPC-CC) in 1999, and much research followed. Further analysis has shown that the decoding thresholds for the LDPC-CC codes are not only better than the belief propagation decoding thresholds of corresponding regular LDPC block codes, but also even very close to the maximum likelihood (ML) decoding thresholds.
This invention includes a simple method to construct a set of SC-LDPC codes with different code rates that is easily ported to hardware via an FPGA-ready configuration.
Applications:
- 5G technologies
- Optical communications
- Information reconciliation
Advantages:
- Simple
- Superior performance
