Course: 2nd semester 2020/2021
Areas: Mobile communications, 4G, 5G, crowdsourced data, radio spectrum
Supervisors: Luis Mendo y Zoraida Frías
Even though the amount of spectrum for mobile services has increased significantly in recent years following the exponential growth in the mobile traffic demand, radio spectrum remains a limited and scarce resource. With the next generation of mobile technologies, 5G, the historical competition for spectrum resources between mobile and other systems will expectedly increase, as new stakeholders in the mobile ecosystem are claiming dedicated spectrum for the different vertical use cases.
This rivalry over spectrum resources will require the development of new technologies to optimize the use of radio spectrum in an increasingly complex and distributed setting. Predicting the demand for spectrum or the network load in real-time is one of the building blocks towards more efficient spectrum sharing mechanisms.
In this context, we aim to explore the potential of a method to measure network load in a distributed way and the capabilities of machine learning techniques to accurately predict network load using real crowdsourced mobile measurements.
To this aim, we will first synthetically generate mobile measurements that reproduce real crowdsourced data to test the sensitivity of the measurement methods to diverse circumstances, including potential biases and incompleteness of the crowdsourced data.
In this Master Thesis, the student will develop a generator of synthetic mobile measurements for existing 4G and future 5G networks. The ultimate objective is to generate diverse datasets of mobile measurements to test the sensitivity of the methods to measure network load under a known and controlled environment.
The generator will be developed in Matlab.
Application to a scholarship is possible.
- Matlab programming skills
- Knowledge of mobile networks’ standards