Ανάπτυξη Αλγορίθμου Για Την Τυφλή Εκτίμηση Του Σηματοθορυβικού Λόγου Σε Δυναμικά Οπτικά Δίκτυα

Abstract

As we are approaching the physical capacity limit of standard optical fiber, we are in a great demand of achieving high utilization-efficiency of the available resources in order to accommodate the geometrical increasing telecommunication traffic. The static nature of the conventional networks makes current networks inappropriate for satisfying future consumers’ needs and expectations. During the last decades, new, flexible optical network architectures have been proposed to overcome the inabilities of the static ones. The main advantage of these networks is the capability of adapting their operation on the link’s varying existing physical and traffic conditions. The most important procedure that takes place in these networks is the optical performance monitoring (OPM) with witch the network controller is provided with the essential information about system’s current state, in order to dynamically allocate the available resources and therefore, to maximize the utilization-efficiency. In this thesis, we examine optical networks which use coherent detection technology for the transmission schemes. The available modulation formats of the networks are 4-QAM (QPSK), 16-QAM, 64-QAM and 256-QAM. A small reference in digital multilevel modulation formats is made. Moreover, an algorithm based on digital signal processing is proposed to reliably estimate the existing signal to noise ratio (SNR) of the link. In addition, we test the performance of various dynamic networks for different transmission parameters. For the OPM of these networks the proposed algorithm is used. Statistical results have been added for the evaluation of the introduced method and it is proven that the estimations of the algorithm are very accurate.