Comparative study for identification of multiple alarms in telecommunication networks


Abstract: Telecommunication networks consist of communication units interconnected physically or by means of protocols in order to provide basic services like data, voice, or image transfers. In this study, a modeling frame for network units and their links in a topological frame is presented based on a real mobile communication network named TASMUS (TAktik Saha MUharebe Sistemi - Tactical Field Combat System). Alarm handling is one of the most critical features required in communication networks. Based on simulated single alarm and multiple (double) alarm scenarios, known powerful alarm estimation approaches, namely the coding method, neural networks, and knowledge-based systems, have been studied to assess their capabilities for identifying multiple faults that might occur simultaneously in real time. They have also been compared in order to evaluate the performance of alarms under different noise levels for specific TASMUS networks.

Keywords: Telecommunications network, network management, neural networks, alarm correlation, fault identification, coding method, minimal distance decoder

Full Text: PDF