Fiber optic link lifetime forecast using mivar AI technologies

The paper describes the integration of mivar artificial intelligence (AI) technologies into methods for predicting the lifetime of fiber optic links. A hybrid model is proposed on the basis of optical fiber fracture models and mivar inference algorithms, which makes it possible to increase forecast precision by taking into account dynamic forces, temperature effects and operation data. Presented calculations demonstrate reduction of error to 0.5%% compared to classical approaches. The advantages of the proposed approach over traditional methods and the prospects for its implementation in the telecommunications industry are also discussed.