Performance Monitoring for Live Systems with Soft FEC and Multilevel Modulation

TL;DR

This work tackles live performance monitoring in coherent optical systems carrying soft FEC and multilevel modulation by introducing a blind ASI estimation method. By storing a histogram of L-values prior to FEC decoding and fitting it to discretized Gaussian candidates, the approach estimates the asymmetrical information $\text{ASI}$ without knowledge of transmitted bits, with a by-product estimate of $\hat{Q}_{st,L}$. Experimental validation across eight modulation/shaping formats shows ASI estimation errors typically below a few percent and around 0.5 dB SNR equivalence for binary channels in the relevant regime, while requiring only modest additional DSP storage. The method enables practical, real-time performance monitoring of live traffic, reducing the gap between modern information-theoretic metrics and industry-grade live-system diagnostics.

Abstract

Performance monitoring is an essential function for margin measurements in live systems. Historically, system budgets have been described by the Q-factor converted from the bit error rate (BER) under binary modulation and direct detection. The introduction of hard-decision forward error correction (FEC) did not change this. In recent years technologies have changed significantly to comprise coherent detection, multilevel modulation and soft FEC. In such advanced systems, different metrics such as (nomalized) generalized mutual information (GMI/NGMI) and asymmetric information (ASI) are regarded as being more reliable. On the other hand, Q budgets are still useful because pre-FEC BER monitoring is established in industry for live system monitoring. The pre-FEC BER is easily estimated from available information of the number of flipped bits in the FEC decoding, which does not require knowledge of the transmitted bits that are unknown in live systems. Therefore, the use of metrics like GMI/NGMI/ASI for performance monitoring has not been possible in live systems. However, in this work we propose a blind soft-performance estimation method. Based on a histogram of log-likelihood-values without the knowledge of the transmitted bits, we show how the ASI can be estimated. We examined the proposed method experimentally for 16 and 64-ary quadrature amplitude modulation (QAM) and probabilistically shaped 16, 64, and 256-QAM in recirculating loop experiments. We see a relative error of 3.6%, which corresponds to around 0.5 dB signal-to-noise ratio difference for binary modulation, in the regime where the ASI is larger than the assumed FEC threshold. For this proposed method, the digital signal processing circuitry requires only a minimal additional function of storing the L-value histograms before the soft-decision FEC decoder.

Figures (9)

• Figure 1: System models: (a) binary modulation without FEC, (b) binary modulation with FEC, (c) multilevel modulation with FEC, and (d) multilevel modulation with FEC and PS. The functions M, $\mathrm{C}_{\mathrm{F}}$, $\Pi$, and $\mathrm{C}_{\mathrm{S}}$ are symbol mapping, FEC encoding, bit-interleaving, and PS encoding at the transmitter. The functions at the receiver are their inverse operations.
• Figure 2: Typical computation of information-theoretic performance metrics (green) and functions for performance monitoring of live systems (orange and yellow) around the FEC decoding in DSP.
• Figure 3: Exemplified histogram of L-values for PS-$64$-QAM having $\mathbb{H}(\boldsymbol{B})$ of $4.1\,\mathrm{bpcu}$ and the SNR of $10\,\mathrm{dB}$. The pmfs are as a function of (a) symmetrized a posteriori L-value $L_{\mathrm{a}}$ and (b) absolute value $|L_{\mathrm{a}}|$, where $\Delta l =1/13$. Without the knowledge of the transmitted bits, we know the histogram $P_{|L_{\text{a}}|}(l)$ in (b). In the proposed method, we compare $P_{|L_{\text{a}}|}(l)$ with candidate histograms $P_{|G_k^{\mathrm{t}}|}$ for $k=1,\ldots K$, and choose the best $k$, i.e., $\hat{k}$ in \ref{['eq:full_search']}.
• Figure 4: Performance of ASI estimation for different $n_\mathrm{bin}$ values for PS-$64$-QAM having a symbol entropy $\mathbb{H}(\boldsymbol{B})$ of $4.1\,\mathrm{bpcu}$ over the Gaussian channel.
• Figure 5: Experimental setup. Insets are recovered constellations for PS-64-QAM in the case of $\mathbb{H}(\boldsymbol{B})=5.7\,\mathrm{bpcu}$ after $5$ roundtrips (top) and $\mathbb{H}(\boldsymbol{B})=4.6\,\mathrm{bpcu}$ after $13$ roundtrips (bottom).