Modeling ultra wideband coherent transceivers and ROADMs for open networking applications

Optical networks deploying high capacity WDM optical data transport are becoming a fundamental infrastructure in modern society to support the ever increasing Internet traffic demand. To achieve an efficient transmission control, virtualization down to the physical layer of devices involved in the whole communication system is required to accurately and realistically simulate the behavior of the network, in particular of a software defined network. For this to happen, modeling and appropriately characterizing all network elements involved in the optical fiber transmission is a fundamental request, to enable software system management and a correct interaction between the different components, possibly multivendor. This step is fundamental for the creation of an open optical network, allowing a cost reduction and a smarter management of the entire communication system. At the same time, a rapid growth of the application market for photonic integrated circuits should be considered, which allows high bandwidth communications and good interaction with classic integrated electronic circuits. For this reason, nowadays more easily integrated technologies are now preferred to classic discrete optical components, usually in lithium niobate, using silicon or indium phosphide. Key components in such a framework are optical I/Q modulators, that are the core of modern transceivers, and reconfigurable add and drop optical multiplexers (ROADMs). So it is specifically required to include such components and their evolution in the softwarization of the optical physical layer. Thanks to the availability of ROADMs and of a prototype of an integrated ultra-wideband IQ modulator in indium phosphide, both provided by Lumentum Company, and thanks to the collaboration with LINKS Laboratories and with the Telecom Infra Project consortium, this study focuses on the characterization of the ROADM, already available as a commercial product, and on the development of a first physical model of the modulator, for the creation of its ”digital twin”, fundamental for the abstraction of the physical layer of a disaggregated open optical network. This thesis work follows, for this reason, two different, complementary approaches, in order to model and characterize the two core components of the network. The thesis goal is to obtain a realistic model for the IQ modulator, starting from a theoretical model, and obtaining simulation results inserting non-linearities of the device and making it working with two different baud-rates (96 and 128 Gbaud), using two different modulation formats (QPSK and 16-QAM), in order to reach the bit-rate goal of 600 and 800 Gbps, obtaining realistic results through inserting the modulator model in a realistic, complete simulator already present in laboratory, called Split Step Fourier Method (SSFM) Simulator. On the other hand, the characterization of the commercial ROADMs by Lumentum is developed in parallel in order to make possible a good network orchestration for further improvements and for future, realistic experiments for a good abstraction and control of the transmission, collocating them in a software defined context.