A Method for Predicting the Usability of Optical Cable Based on Combination Model

Abstract

A combination model predicting the optical cable loss is built by combining three different models, as the grey model, the 3-order exponential smoothing model and the neural network model, by using weighting coefficient. The simulation result and the prediction error calculation show that the combination model has high accuracy to predict the optical cable loss. And then, based on the combination prediction model, a method for evaluating the cable's usability is proposed. The method contains two important parameters as the maximum length of the relay section and the range of allowable loss in the cable line. To assess whether the cable line is available these two evaluation indices must be synthesized to consider. Only when both of two standards are met, the cable line's availability is strong. If one of the standards does not meet the requirements of the network construction and equipment upgrade, technical measures can be taken to enhance the cable performance and improve the availability of optical cable.