Abstract
We analyze the achievable information rates (AIRs) for coded modulation schemes with QAM constellations with both bit-wise and symbol-wise decoders, corresponding to the case where a binary code is used in combination with a higher-order modulation using the bit-interleaved coded modulation (BICM) paradigm and to the case where a nonbinary code over a field matched to the constellation size is used, respectively. In particular, we consider hard decision decoding, which is the preferable option for fiber-optic communication systems where decoding complexity is a concern. Recently, Liga et al. analyzed the AIRs for bit-wise and symbol-wise decoders considering what the authors called "hard decision decoder" which, however, exploits soft information of the transition probabilities of discrete-input discrete-output channel resulting from the hard detection. As such, the complexity of the decoder is essentially the same as the complexity of a soft decision decoder. In this paper, we analyze instead the AIRs for the standard hard decision decoder, commonly used in practice, where the decoding is based on the Hamming distance metric. We show that if standard hard decision decoding is used, bit-wise decoders yield significantly higher AIRs than symbol-wise decoders. As a result, contrary to the conclusion by Liga et al., binary decoders together with the BICM paradigm are preferable for spectrally-efficient fiber-optic systems. We also design binary and nonbinary staircase codes and show that, in agreement with the AIRs, binary codes yield better performance.
Introduction
In high-speed coherent fiber-optic systems, Forward Error Correction (FEC) is essential to maintain data integrity. While Soft-Decision Decoding (SDD) offers superior performance, its high power consumption and complexity make Hard-Decision Decoding (HDD) the industry standard for multi-terabit links. A recent debate has emerged regarding whether binary codes (bit-wise) or nonbinary codes (symbol-wise) are more effective for HDD. Previous research suggested nonbinary codes were superior, but this conclusion relied on a "channel-aware" decoder that utilizes soft information. This paper re-evaluates this comparison using standard HDD based on the Hamming distance metric, which is the actual implementation used in hardware.
Methodology
The study employs a rigorous information-theoretic framework to compare decoding strategies:
- Generalized Mutual Information (GMI): AIRs are derived for bit-wise and symbol-wise decoders under a mismatched decoding framework, correctly modeling standard HDD decoders that treat the channel as symmetric.
- Density Evolution (DE): New DE equations were developed for nonbinary staircase codes to predict performance in the limit of infinite code length.
- System Simulation: Findings were validated using the Split-Step Fourier Method (SSFM) for single-channel links and the Gaussian Noise (GN) model for multi-channel Wavelength Division Multiplexing (WDM) systems.
- Code Design: Binary and nonbinary staircase codes were specifically designed and compared to validate theoretical AIR predictions.
Theoretical Insights: The Advantage of Bit-Wise Decoding
The primary finding is that when restricted to standard HDD using Hamming distance, bit-wise decoders (binary) significantly outperform symbol-wise (nonbinary) decoders. This contradicts previous findings that assumed decoders had access to transition probabilities (soft information).
As shown in Figure 3, the bit-wise GMI ($I_{HDD-BW}^{gmi}$) is consistently higher than the symbol-wise GMI ($I_{HDD-SW}^{gmi}$). Bit-Interleaved Coded Modulation (BICM) effectively averages the error probabilities across different bit positions in high-order constellations like 64-QAM, creating a symmetric channel that binary HDD decoders are well-optimized to handle.
Numerical Results and System Performance
The theoretical gains translate into significant practical advantages in both SNR requirements and transmission reach.
- SNR Gains: For 64-QAM at a spectral efficiency of 3 bits/symbol, the bit-wise decoder requires approximately 2 dB less SNR than the symbol-wise decoder to achieve the same performance.
- Staircase Code Performance: Binary staircase codes outperformed nonbinary versions by 0.6 dB to 1.11 dB across various overheads, confirming the AIR analysis.
- Optical Reach: In WDM simulations, binary coding enabled much longer transmission distances.
Figure 5 illustrates the reach for different modulation formats. For instance, using binary codes with bit-wise decoding resulted in a reach of 701 km for 16-QAM, 590 km for 64-QAM, and 311 km for 256-QAM.
Conclusion and Future Work
This research settles the debate by demonstrating that binary codes combined with BICM are the superior choice for standard HDD in coherent fiber-optic systems. Nonbinary codes are not only more complex to implement but also perform worse under the constraints of real-world hardware. Future work could explore probabilistic shaping to further enhance HDD performance and investigate the impact of channel memory on the decoding process.
Summary updated