Diagnosing the Technical Condition of Heterogeneous Networks of a Spatially Distributed Monitoring System Based on Neural Network Analysis of Telemetry Data

The article describes a neural network approach to the analysis of telemetry data in the interests of diagnosing the technical condition of telecommunications equipment heterogeneous monitoring networks. A description of the operation of a recurrent neural network is given, which provides a solution to the problem of predicting failures based on telemetry data. The results of assessing the effectiveness of neural network training were obtained using the example of temperature data from the control elements of the monitoring system.