Detecting changes in UERC switches : A sequence analysis of UERC switches in  a mobile network

Abstract: This thesis investigates the possibility to analyse a mobile network with sequences of UERC switches specific to each user equipment. An UERC is essentially a channel that carries information and a user equipment connects to different UERCs depending on whether they want to talk and/or send data with different qualities. As a major player in the mobile technology industry, Ericsson strives to optimise the use of the UERCs and are looking for an automated way to detect changes. The first task was to identify and retrieve the required events from the network log files in order to create the UERC sequences. As a way to test the thesis assumption, and give a proof of concept, two different data sets were analysed were changes had been made to the network settings that should have affected the UERC sequences. With the use of n-grams, Markov chains and Bayesian Estimation testing, the changes could be identified and the thesis assumption could be confirmed - UERC sequences provides a possible way of analysing a mobile network.