Frequency Dependent ENoB Requirements for 400G/600G/800G Optical Links

• Published in: Journal of lightwave technology Vol. 38; no. 18; pp. 5008 - 5016
• Main Authors: Varughese, Siddharth, Lippiatt, Daniel, Tibuleac, Sorin, Ralph, Stephen E.
• Format: Journal Article
• Language: English
• Published: New York IEEE 15.09.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Analog to digital conversion; Analog to digital converters; Analog-digital conversion; Bandwidth; Broadband; Communications systems; Computational modeling; Computer simulation; Digital to analog conversion; Digital to analog converters; digital-analog conversion; effective number of bits; electronic converters; Frequency dependence; Machine learning; Optical communication; Optical fiber communication; Optical transmitters; Quantization (signal); Regression analysis; Signal quality; Signal resolution; Timing jitter
• ISSN: 0733-8724; 1558-2213
• DOI: 10.1109/JLT.2020.3000177

Electronic converters (ECs), such as Digital-to-Analog Converters (DACs) and Analog-to-Digital Converters (ADCs), have become essential components in high-speed optical communication systems. However, with increasing operating symbol rates and modulation formats, frequency dependent ENoB such as those demonstrated by wideband ECs begin to impair signal quality significantly. Understanding these impairments is crucial for future optical link planning and deployment. Here, we present a novel ADC model that can simulate frequency dependent ENoB in optical links. The model is both theoretically and experimentally validated. Using such models, we demonstrate the effects of wideband ECs on 64 GBaud DP-16QAM (400G), 64 GBaud DP-64QAM (600G), and 96 GBaud DP-32QAM (800G). Furthermore, we present a machine learning based regression that can analytically obtain EC induced penalties for these investigated systems.