Abstract
Big data analytics and machine learning applications are often used to detect and classify anomalous behaviour in telecom network measurement data. The accuracy of findings during the analysis phase greatly depends on the quality of the training dataset. If the training dataset contains data from network elements (NEs) with high number of failures and high failure rates, such behaviour will be assumed as normal. As a result, the analysis phase will fail to detect NEs with such behaviour. Effective post-processing techniques are needed to analyse the anomalies, to determine the different kinds of anomalies, as well as their relevance in real-world scenarios. Manual post-processing of anomalies detected in an Anomaly Detection experiment is a cumbersome task, and ways to automate this process are not much researched upon. There exists no universally accepted method for effective classification of anomalous behaviour. High failure ratios have traditionally been considered as signs of faults in NEs. Operators use well-known key performance indicators (KPIs) such as drop call ratio and handover failure ratio to identify misbehaving NEs. The main problem with these KPIs based on failure ratios is their unstable nature. This paper proposes a method of measuring the significance of failures. The usage of this method is proposed in two stages of anomaly detection: training set filtering (pre-processing stage) and classification of anomalies (post-processing stage) using an automated process.
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.References
Kumpulainen P (2014) Anomaly detection for communication network monitoring applications. Doctoral Thesis in Science and Technology, Tampere University of Technology, Tampere
Kumpulainen P, Hätönen K (2008) Anomaly detection algorithm test bench for mobile network management. In: MathWorks/MATLAB user conference Nordic. The MathWorks conference proceedings
Chernogorov F (2010) Detection of sleeping cells in long term evolution mobile networks. Master’s Thesis in Mobile Technology, University of Jyväskylä, Jyväskylä
Kumpulainen P, Kylväjä M, Hätönen K (2009) Importance of scaling in unsupervised distance-based anomaly detection. In: Proceedings of IMEKO XIX World Congress, fundamental and applied metrology. Lisbon, Portugal, pp 2411–2416
Kohonen T (1997) Self-organizing maps. Springer, Berlin
Laiho J, Raivio K, Lehtimäki P, Hätönen K, Simula O (2005) Advanced analysis methods for 3G cellular network. IEEE Trans Wirel Commun 4(3):930–942
Kumpulainen P, Hätönen K (2008) Local anomaly detection for mobile network monitoring. Inf Sci 178(20):3840–3859
Yin H (2008) The self-organizing maps: background, theories, extensions and applications. Comput Intell Compend Stud Comput Intell 115:715–762
Suutarinen J (1994) Performance measurements of GSM base station system. Thesis (Lic.Tech.), Tampere University of Technology, Tampere
Hätönen K, Kumpulainen P, Vehviläinen P (2003) Pre and post-processing for mobile network performance data. In: Finnish Society of Automation, Helsinki, Finland, September
Anonymous. k-means clustering. Mathworks. http://se.mathworks.com/help/stats/k-means-clustering.html. Accessed 20 Dec 2014
Höglund AJ, Hätönen K, Sorvari AS (2000) A computer host-based user anomaly detection system using the self-organizing map. IEEE-INNS-ENNS Int Joint Conf Neural Netw (IJCNN) 5:411–416
Kylväjä M, Hätönen K, Kumpulainen P, Laiho J, Lehtimäki P, Raivio K, Vehviläinen P (2004) Trial report on self-organizing map based analysis tool for radio networks. Veh Technol Conf 4:2365–2369
Anonymous. Serve atOnce Traffica. Nokia Solutions and Networks Oy. http://networks.nokia.com/portfolio/products/customer-experience-management/serve-atonce-traffica. Accessed 16 Dec 2014
Anonymous. MATLAB—the language of technical computing. Mathworks. http://se.mathworks.com/products/matlab/. Accessed 12 June 2016
Ward JJH (1963) Hierarchical grouping to optimize an objective function. J Am Stat Assoc 58(301):236–244
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Jerome, R.B., Hätönen, K. Anomaly detection and classification using a metric for determining the significance of failures. Neural Comput & Applic 28, 1265–1275 (2017). https://doi.org/10.1007/s00521-016-2570-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-016-2570-7