Machine Learning for Classifying Cellular Traffic

Abstract

Today’s cellular network is ever growing, making the need for a mechanism that can identify overloads greater each day. In this report a design science research is conducted showcasing the possibilities to use the classification machine learning algorithm naive Bayes to identify signaling overloads in a cellular network node. The research shows that naive Bayes can be used to successfully identify the greater majority of the possible overloads that could occur in a cellular node.