Accurate model to predict performance of coherent optical transponder

Abstract

Coherent optical transponder is ubiquitous and dominant in long-haul optical network. Bit error rate (BER) versus optical signal to noise ratio (OSNR) determines the transmission distance for coherent optical transponder. However, a complicated setup is needed for this measurement, which limits this measurement to laboratory environment. We have developed an accurate model to predict BER versus OSNR at various receiver optical power (ROP) under assumption of additive white Gaussian noise (AWGN). The model has three parameters, which are related to noise floor, filter mismatching, and OSNR value without noise loading. We determined the first two parameters through curve fitting of BER vs. ROP curve. We determined the third parameters through design verification test (DVT). We validated the model over 100 channels within extended C band. Furthermore, we expanded the model to high modulation format 16- ary quadrature amplitude modulation (QAM). We investigated the influence of high baud rate, like 45G, 56G, 64G and 86G. The model works well for both high modulation format and high baud rate. The influence of baud rate on the fitting parameters are discussed. Since one can measure BER versus ROP using built-in components of coherent optical transponder, BER versus OSNR can be monitored during in-field deployment based on this accurate model. In addition, one can monitor OSNR based on BER reversely. No extra hardware or DSP processing algorithm is needed for this OSNR monitoring scheme. The monitoring accuracy is further improved with consideration of chromatic dispersion (CD), polarization mode dispersion (PMD) and nonlinearity impairment.