Study On Intelligent Monitoring Of Optical Signal Parameters In Elastic Optical Network

With the continuous development of science and technology,large broadband services have sprung up.Especially in mobile communication technology,big data and video conferencing,etc,the variety of user requirements in the network has led to the explosive growth of bandwidth requirements.The capacity of optical communication systems continues to increase,and the architecture of optical networks continues to become more complex,transparent and dynamic in nature.The elastic optical network has attracted great attention due to its more flexible spectrum allocation mechanism than the wavelength division multiplexing system.Due to the inherent dynamic characteristics of the transmitting end of the elastic optical network,the system parameters will change dynamically with the changes of optical fiber link conditions and user requirements.In addition,mode-division multiplexing transmission systems based on few-mode fibers have attracted considerable attention,which makes the optical network architecture more complex.In this case,in order to adopt an appropriate damage compensation algorithm and allocate network resources more efficiently,Optical Performance Monitoring（OPM）is considered to be a common requirement for future elastic optical networks based on single-mode and few-mode.When performing OPM,Modulation Format（MF）and Optical Signal to Noise Ratio（OSNR）,as key parameters for optical performance monitoring,are the focus of research.This paper focuses on the key technologies of optical performance monitoring in elastic optical network systems.The main research contents are summarized as follows:1.We have proposed an MFI scheme with a low computational complexity,which combines an improved particle swarm optimization（I-PSO）clustering algorithm with two-dimensional（2D）Stokes plane.The main idea of I-PSO is to add a new field of view on each particle and limit each particle to only communicate with its neighbor particles,so as to realize the correct judgment of the number of multiple clusters（local extrema）on the density images of s₂-s₃ plane.The effectiveness has been verified by 28 GBaud polarization division multiplexing（PDM）-BPSK/QPSK/8QAM/16QAM/32QAM/64QAM simulation EON systems and 28 GBaud PDM-QPSK/PDM-8QAM/PDM-32QAM proof-of-concept transmission experiments.The results show that using this MFI scheme,the minimum OSNR values to achieve 100%MFI success rate are all equal to or lower than those of the corresponding 7%forward error correction（FEC）thresholds.At the same time,the MFI scheme also obtains good tolerances to residual chromatic dispersion（CD）and differential group delay（DGD）.Besides that,the proposed scheme achieves 100%MFI success rate within a maximum launch power range of-2～+6 d Bm.More importantly,its computational complexity can be denoted as O（N）.2.A multi-task neural network based MFI and OSNR joint monitoring scheme for mode division multiplexing（MDM）system is proposed.Six spatial modes are considered,including LP₀₁,LP_11a,LP_11b,LP_21a,LP_21b,LP₀₂,and six common modulation formats such as PDM-BPSK/QPSK/8QAM/16QAM/32QAM/64QAM are selected.The constellation diagram is sent to the multi-task neural network as the input feature,and the effectiveness of the proposed scheme is verified by building a VPI simulation system.For the classification task of modulation format recognition,the training set accuracy rates of the three spatial patterns LP₀₁,LP_11a,and LP_21a reached 99%after 116,122,and 110 iterations,respectively.The corresponding test set accuracy reaches 99%after 146,148 and 151 iterations,respectively.For the regression task of OSNR monitoring,MAEs of the three spatial mode training sets of LP₀₁,LP_11a,and LP_21a reached 0.32d B,0.33d B,and 0.28d B after 146,152,and 163 iterations,respectively;MAEs of corresponding test sets reached 0.32d B,0.35d B and 0.30d B after 151,149 and 181 iterations,respectively.The results show that the multi-task learning-based joint monitoring scheme has good generalization ability for both training set and validation set,and achieves good modulation format identification and OSNR monitoring results.