The main objective of this project is to develop a methodology to evaluate the performance and optimize the QOS of the LTE mobile network. This methodology allows to detect network failures and to minimize them from dedicated measurements via the DT drive test measurement chain driven by the Nemo Outdoor software. In our project, we focused on improving the quality of service in the three main parameters: coverage, signal quality and throughput. The choice of the solution is validated by the interpretations of different simulations provided by the optimization tool Atoll. Thus, our project aims at implementing machine learning algorithms to predict the quality of service of an LTE network.