Telecom service providers are faced with an overwhelming flow of alarms, which makes good alarm classification and prioritisation very important. This paper first provides statistical analysis of data collected from a real-world alarm flow and then presents a quantitative characterisation of the alarm situation. Using data from the trouble ticketing system as a reference, we examine the relationship between mechanical classification of alarms and the human perception of them. Using this knowledge of alarm flow properties and trouble ticketing information, we suggest a neural network-based approach for alarm classification. Tests using live data show that our prototype assigns the same severity as a human expert in 50% of all cases, compared to 17% for a naïve approach.Keywords: communication system operations and management; neural network applications; alarm systems.Reference to this paper should be made as follows: Wallin, S., Leijon, V. and Landén, L. (2009) 'Statistical analysis and prioritisation of alarms in mobile networks', Int.

Abstract-Telecom Service Providers are faced with an overwhelming flow of alarms. Network administrators need to judge which alarms to resolve in order to maintain the service quality. The problem is that it is hard to pick the most important alarms. Which alarms have the highest priority? A solution that automatically assigns priorities to alarms would increase the efficiency of Network Management Centers. We have prototyped a solution that uses neural networks to assign alarm priority. The neural network learns from network administrators by using the manually assigned priorities in trouble-tickets. Our tests are based on live-data from a large mobile service provider and we show that neural networks can learn to assign relevant priorities to 75% of the alarms.Index Terms-Communication system operations and management, Neural network applications, Alarm Systems.