Data Science Applied to 4G and 5G Telecom Networks: A Case Study Using Unsupervised Clustering Algorithms to Identify Event Patterns in Antennas.

Recent advances in the Telecommunications industry have led to positive advances in terms of obtaining valuable data, resulting from the availability of next generation networks (5G) and more and more devices connected to mobile networks. Data from mobile networks can be useful to improve companies’ strategies. Since there are many potential benefits from obtaining the data mentioned, the challenge is to obtain the information efficiently and present them in a clear and understandable way. This work presents a methodological approach based on unsupervised machine learning to handle the analysis of errors registered in the activity logs of the eNodeBs antennas of a Telecommunications company.

During the tests, the model proved to be effective in grouping similar data and enabling the elaboration of meaningful graphical visualizations.