Detección de degradaciones en redes móviles basado en máquina de estados y umbrales de Otsu

Abstract

Troubleshooting degradations in a mobile network is mainly a manual activity. Self-Healing (SH) appeared to automate such functions, using techniques such as Data Analytics with network performance indicators. Along this line, this work proposes an algorithm to detect degradations, analysing network metrics. The algorithm is based on a state machine that follows the status of the network over time series of indicators. This algorithm only requieres the specification of few parameters; some of them calculated using Otsu’s thresholding. Finally, the performance of the algorithm has been checked using data obtained from a LTE microcell.