5G traffic forecasting through the use of social networks in a realistic scenario

The research work to be done in the master thesis is grouped in two parts. In the first one a realistic 5G scenario will be defined by using real data from an already existing 4G network: site information (location, sectorization, etc.), UL/DL traffic information per cell during a time period of around three months, and number of PRBs assigned/used per cell. With this data a C-RAN network with NFV and considering the use of slicing and MEC for specific services and locations, will be designed. The second part plans to use the large amount of data that can be extracted onlinSocial networks have marked a before and after in data management. The exponential increase in the amount of information online in recent years is undeniable. New tools and systems are needed more frequently to optimize the management of mobile networks in the face of massive events (vaccination days, protests, parades, etc). In this way, the present work proposes the design of a realistic 5G scenario based on real data from an existing 4G network and with these data design a C-RAN with NFV. Additionally, it is proposed to use the amount of information that can be extracted from social networks about massive events in a city and to link this information with the network elements and their performance. With this, it is possible to forecast the presence of crowds in specific locations and thus relocate resources to the cells that will meet that demand