I am a student of the Master in Industry 4.0 in the Univerity of Pau and Pays de l'Adour, (UPPA) and a student of the Master in Computer Science in the Autonomous University of Yucatan (UADY). I have the B. Sc. degree of Applied Mathematics and Computation of the National Autonomous University of Mexico (UNAM). I have knowledge in the areas of networking, cloud computing, machine learning, and database management.
I enjoy most of my time being indoors. I love reading fiction, fantasy, and books about history, and theology.
As thew Internet grows, so does the need of implementing quality controls. Quality of Service (QoS) provides a set of rules and standards that allows the analysis of quality that a service throughout a network has. One of the means of assuring QoS is through the classification of information packages. A classifying method that has demonstrated to have a high level of accuracy is classification through Machine Learning.
Machine Learning requires ground data to learn a behavior. Sadly, there's not enough current data to train a Machine Learning Algorithm (MLA) without implementing other tools such as Deep Packet Inspection (DPI) tools. In this project, we propose a cloud architecture that automatically generates the required data to train different MLAs through virtual machines. The proposed architecture will provide each virtual machine a behavior reminiscent to that of the human, such as web browsing, online streaming, and so on. The internet traffic generated by the virtual machines pass though a gateway which sniffes the activity and stores it in a database, which will be used to train the MLAs.
The following is a generalization of the architecture that shows how the labeling process works in conjunction with the sniffing process to generate a database for the MLAs
In addition, we present the operational analysis and the system analysis, to show the implementation of different scenarios in the architecture.