Research On Perceptual Transmission Technology Of Cognitive Optical Network

In recent years,driven by smart phones,cloud computing,and the Internet of Things,network traffic has grown rapidly,and there is an urgent need for large-capacity,highefficiency bearers and intelligently managed networks.However,the traditional WDM optical network has low efficiency in spectrum resource utilization and poor dynamic flexibility,which cannot cope with the increasingly complex dynamic optical network environment.Cognitive optical network is an intelligent network with autonomous awareness and control.The ability of self-learning and decision making enables cognitive optical networks to have greater flexibility and self-healing capabilities,and it can make rational use of spectrum resources to meet the needs of future optical network services.Therefore,this thesis combines artificial intelligence technology to deeply study the sensing method and transmission technology of cognitive optical network,so that it can realize intelligent sensing and control while improving spectrum utilization.For the application of cognitive optical networks,this paper first analyzes the architecture of cognitive optical networks and summarizes the effective cognitive models and parameters perceived by optical networks based on their characteristics.Then,based on the back propagation neural network,the thesis proposes an optical signal to noise ratio(OSNR)estimation technique,which uses the optical path with routing and wavelength information as the input of the neural network,and the OSNR of the optical path as the output of the neural network to realize the optical network link OSNR estimate.In the research,the thesis analyzes the effects of neural network iterations,training data and number of neurons on OSNR estimation by numerical simulation.In addition,considering the phenomenon that the optical signal quality is degraded by the EDFA amplification noise and the inter-channel nonlinear interference noise(NLI)during transmission,this thesis also focuses on the influence of NLI noise on OSNR estimation.In addition,in order to improve the transmission capacity within a limited spectrum bandwidth,this thesis proposes a method for spectrum sharing of primary and secondary signals.The method can add a suitable secondary signal according to the OSNR margin of the primary signal and detect the secondary signal based on the serial interference cancellation technique at the receiving end.In the thesis,the feasibility of the simulation is verified by numerical simulation.The effects of OSNR margin,signal power and transmission rate on the whole transmission system are studied respectively.The maximum power and transmission rate that can be set from the signal under different OSNR margins are plotted.At the same time,this thesis attempts the low-cost binary onoff keying(OOK)modulation format of the secondary signal and compares the performance difference caused by different detection methods.The simulation results show that the SIC-based coherent detection performance is better than the amplitude detection.