After several decades where research in mobile communications have mainly focused on improving user data rates, we are now at a turning point: while future mobile applications will be elastic with respect to the available capacity, they will be extremely sensitive to latency and reliability. Examples are mixed/virtual reality, vehicular and drone networks, and the Internet of Things.
To efficiently support these new requirements, as well as to cope with an ever increasing number of heterogeneous wireless devices, some disruptive solutions are required for future wireless networks:
1) flexible spectrum access strategies, redefining the concept of licensed spectrum, and fully supporting cognitive-like radio solutions;
2) exploiting network densification by using cooperative multi-Access Point (AP) solutions able to use interference in a constructive way, and
3) fast network adaptation to find optimal (or close to) solutions in very dynamic scenarios.
To advance towards the aforementioned vision, our research targets:
1) New spectrum access solutions: Design of novel strategies and protocols for maximizing the reuse of the spectrum resources in multiple dimensions (space, frequency and time) aiming to guarantee both latency and reliability requirements while minimizing the loss on network capacity. Special interest is on the simultaneous use of multiple-bands.
2) Network cooperation: The presence of multiple overlapping access points allows to consider inter-AP cooperation to improve the network operation. Thus, we aim to explore which are the theoretical gains of network cooperation, in which conditions a group of overlapping APs should cooperate (as a tradeoff between performance gains & overheads), and to design efficient resource allocation and user scheduling strategies.
3) Machine Learning: In order to support previous points, decisions about next actions/configurations can be made at the network edge or in the cloud depending on the amount of available information and the maximum tolerable delay. Combining efficiently short-term local decisions based on partial information with much more accurate but mid/long term decisions based on a global view of the network is still an unsolved challenge.
4) Protocol design by Learning: The use of artificial intelligence to build new communication protocols and functionalities able to adapt to every possible situation is a new research field that will completely revolutionize the design of future networks, allowing to reach unexpected levels of efficiency and performance.
Overall, we aim to continue contributing to the development of future wireless networks. We believe that with our current expertise and background on wireless networking and machine learning, we are in a very good position to offer fresh and effective solutions to the open challenges, and significantly push forward the state of the art.