Mobile systems are increasing in number and, in the future, exponential growth is expected with the deployment of new technologies like 5G and Internet of Things. Requirements from those technologies lead to an improvement from the existent techniques to new sophisticated ones. A key role in future developments, which are already applied in research and industry, are Software Defined Networks (SDN) and Network Function Virtualization (NFV). Therefore, we present a solution for mobile edge computing (MEC) using a deep reinforcement learning (DRL) algorithm to optimize and offload tasks in a scenario of a virtual radio access network (VRANs). Final chapters show results obtained from experiments where the learning agent improves its reward through time benefiting the amount of bandwidth used in the network. Finally, a chapter discussing about the conclusions arise with interesting future work which could potentially lead to better results.