A prior-training-free and low-complexity modulation format identification (MFI) scheme 14 based on amplitude histogram distributions is proposed and demonstrated, both numerically and 15 experimentally, for autonomous digital coherent receivers. In the proposed scheme, after having 16
performed power normalization, incoming polarization division multiplexed (PDM) signals are 17 classified into QPSK, 8QAM, 16QAM, 32QAM and 64QAM signals according to their ratios defined 18 according to specific features of their amplitude histograms. The proposed MFI scheme uses ampli- 19 tude information only, thus it is insensitive to carrier phase noise. Furthermore, the proposed 20 scheme does not require any prior information such as optical signal-to-noise ratio (OSNR). The 21 performance of the proposed MFI scheme is numerically verified using 28GBaud PDM-QPSK/- 22 8QAM/-16QAM/-32QAM/-64QAM signals. The numerical simulation results show that the pro- 23
posed scheme can achieve 100% of correct identification rate for all of the five modulation formats 24 when their OSNR values are higher than the thresholds corresponding to the 20% FEC correcting 25 bit error rate (BER) of
2.4 10− .To further explore the effectiveness of the proposed MFI scheme, 26
proof-of-concept experiments in 28GBaud PDM-QPSK/-8QAM/-16QAM, and 21.5GBaud PDM- 27 32QAM transmission systems are also undertaken, which show that the proposed scheme is robust 28 against fiber nonlinearities. To explore the scheme’s feasibility for use in practical transmission sys- 29
tems, the computational complexity analysis of the proposed scheme is conducted, which shows 30 that, compared with relevant MFI scheme, the proposed MFI scheme can significantly reduce the 31 computational complexity