Use of Extreme Value Statistics to Assess the Performance Implications of Cascaded ROADMs

The performance implications of passband impairments and bandwidth narrowing caused by the cascading of optical filters are investigated using a 100 Gb/s dual polarization, quadrature-phase-shift-keyed (DP QPSK) transceiver. The overall responses for cascades of filters are emulated using the combination of a programmable optical filter and variable bandwidth optical filter. The statistical variations in the responses for a cascade are addressed by considering 1000 realizations that result from randomly selected responses for each of the individual filters. To determine the impact on system margins, a methodology based on extreme value statistics is presented. The signal-to-noise ratios (SNRs) that correspond to real-time estimates of the pre-forward error correction bit error ratio (pre-FEC BER) are quantified in terms of the probability that the minimum of
$n$
observations of the SNR is less than a specified value (maximum pre-FEC BER exceeds a specified value). This approach improves the reliability of predicting the impact of cascaded filtering on system performance by removing the uncertainty about the underlying statistical distribution of the SNR.