Edge cloud computing seems to be a key enabler of 5G networks which essentially brings the servers as close to the users as possible. Among all the benefits this tendency can provide, this master thesis focuses on the advantages in terms of reduction of the latency. First of all, an Edge network model that combines this paradigm with Software-defined Networks (SDN) is presented so as to provide an example of a potential production scenario. Then, a videoconference application is chosen as a particular case study of latency-sensitive and bandwidth exhaustive application and the traffic that it generates is inspected. Thanks to this analysis, a methodology to compute the latency can be proposed which is used during the test runs afterwards. Lastly, a testbed analogous to the model previously presented showcases the benefits of this approach. The results prove the improvement in the quality of the videoconference by means of a noticeable reduction of the latency when the servers are on the edge. Moreover, it is demonstrated the feasibility of providing a dynamic environment where the server can be live migrated. For the sake of providing a complete quality overview, the impact of the available bandwidth and packet loss is evaluated as well.