| All-Optical Networks are the next generation optical networks that will provide a transport mechanism to deliver a variety of users' data telecommunications and quality of service requirements reliably. Their desirable features such as transparency to the information transfer, enormous bandwidth make them a natural choice among telecommunications service providers. However, all-optical networks' wide commercial deployments are hampered by their inadequacy to provide a cost effective optical performance monitoring of the user's data and the negotiated quality of service requirements. Many optical monitoring techniques designed to measure the optical signal quality have been proposed in the literature. All of the proposed methods require either the use of monitoring devices by tapping into a portion of the optical fiber link to estimate optical power levels, optical signal to noise ratio, or performing Q-factor measurements using histograms, among other methods. Most of these techniques require the use of a great deal of electronic processing at the device level, or some level of access to the information bit stream in the core network. As a result, these optical monitoring methods hinder transparency to data bit rate, code or modulation format as well as the speed of the information transfer. Additionally, all the currently proposed optical performance monitoring approaches limits to point-to-point physical layer monitoring, ignoring the rest of the networks. In this dissertation, we propose to perform analysis of an optical monitoring methodology that incorporates the user access layer, the physical layer and the network layer on integrated optical meshed network architecture. We investigated the performance parameters that need to be monitored at each layer. We modeled a wavelength routed transparent network, and performed cost optimization at three separate layers; the optical transport layer, the multiplex layer, and the optical channel layer. We then determined the optimum transparency reach that minimizes network equipment cost, node cost, and link cost based on random traffic distributions. We performed physical layer performance simulations of WDM systems and optical meshed-networks to collect statistics on optical performance parameters that most affect the signal quality in the present of noise, distortion, and non-linearity and during network failures. We performed analysis and evaluation of these parameters using physical layer network simulation of WDM systems and optical fiber transmission systems under various optical amplifier noise conditions. Simulation results of our optical performance analysis architecture have shown greater improvement in accuracy in monitoring and measurements of the performance parameters, in addition to being more cost effective compared to other proposed optical performance monitoring approaches. |
