Impact of in-band crosstalk and amplified spontaneous emission noise in differentially phase-shift keying signal

The relative impact of in-band crosstalk and amplified spontaneous emission (ASE) noise on a system's performance has been investigated with a differentially phase-shift keying signal. I have first measured an ASE noise-induced penalty without any in-band crosstalk and then estimated the system's penalty with a simple addition of the measured ASE noise-induced penalty and the calculated in-band crosstalk-induced penalty. Using this approach, the estimated penalty agreed well with a measured system penalty when the Q value was 3 or an optical signal-to-noise ratio (OSNR) of the signal was higher than 30 dB at Q=6. To estimate the system penalty with a low OSNR level at Q=6, an addition of signal-ASE and signal-in-band crosstalk beat noises were added with a weighting factor. Based on this approach, a discrepancy between the estimated and the measured penalties was reduced drastically with an OSNR of 20 dB at Q=6. However, a small discrepancy was still observed even with the weighted addition of two beat noises. Thus, I have confirmed that the effect of in-band crosstalk-in-band crosstalk beat noise should be taken into account for the proper estimation of a system penalty with an OSNR of <30 dB at Q=6.