Estimation of an Abnormal Noise Location Over Multi-Span Optical Fiber Link Based on Nonlinear Fourier Transform

In the next-generation optical fiber networks, in addition to ultra-high speed and high capacity, early recovery from disasters and failures is required to realize highly reliable networks. Recently, various methods for estimating abnormal locations using coherent receivers and digital signal processing have been proposed. This research proposes the application of nonlinear Fourier transform (NFT) for detecting abnormality. Particularly, we explore a method for estimating the abnormal noise location by monitoring the variances and covariance of the discrete eigenvalues in the NFT. The feasibility and characteristics of the proposed method are investigated through numerical simulations. For the details of the proposed method, we investigate the dependency of the eigenvalue configuration of signals, which is used for abnormality detection. Moreover, we confirm that the proposed method remains effective with variations in the noise level and the location of the abnormal noise. Ultimately, the proposed method achieves a kilometer-order estimation accuracy by employing appropriate parameters. In addition, we demonstrate that the proposed method is applicable to the case where the launch optical power is shifted from the optimum or an optical fiber link with a different span length.