“Telecommunications are really at the basis of what is called digital revolution”

We interviewed Carla Chiasserini, Full professor from Politecnico di Torino

07 October 2024

We interviewed Carla Chiasserini, member of the Scientific Council of IMDEA Networks. She graduated (110/110 summa cum laude) in Electronic Engineering from the University of Florence in 1996, and received her Ph.D. in Electronic Engineering and Telecommunications from Politecnico di Torino in 2000. Since then, she has been working at the department of Electronics and Telecommunications Department of Politecnico di Torino, where she is currently a Full Professor. She is also a Research Associate of the Italian National Research Council (CNR).

Her expertise is in wireless networks. Recently, she moved more towards the study of mobile networks and mobile services. “And since many mobile services and mobile applications use machine learning as an essential component, of course, I moved a bit to the machine learning area as well, which is very popular nowadays in our scientific community, but what I try to do has been to identify some aspects that lie on the edge between networking and more fundamental machine learning aspects”.

What are you working on now?

I’m working within the Predict-6G project that’s one of the European projects I’m on working on, which is coordinated by the University Carlos III, and I also work on the CENTRIC project and the ADROIT6G project. They all include investigations about, of course, networks and the use of machine learning to make the network more performing, to improve the network performance and at the same time to reduce the energy impact of machine learning on the ICT systems.

But what actually fascinates me more is to look at the structure of machine learning models as there is a recent trend that is referred to as Dynamic neural networks, where you actually have different blocks that compose the neural network like a jigsaw puzzle made of different pieces and what you can do is to deploy the different pieces on the network nodes in order indeed to match with the nodes resources. You can also choose which blocks of the neural networks, like choosing the pieces essentially of the jigsaw puzzle that you need to implement your application, and at the same time they are suitable for the resources of the mobile network. So, in that sense I hope that my research will contribute indeed to making machine learning more sustainable, because it will allow us to reuse computational networking resources that are already there and to use them more wisely.

What are your research interests?

I think that it will be important to look more deeper into data and the data characteristics, and the data that can be collected or data that can be generated synthetically and try to add this additional dimension to the problem that I was mentioning. Because it’s not just about consuming computational and networking resources in order to support applications based on machine learning but it’s also clearly about managing and collecting data and whenever we collect data some energy is consumed and some resources are consumed and we need to transfer this data from one node to another depending on where the data needs to be processed.

Understanding which data is really important which data can be synthetically generated within the network without the need to collect this data from the nodes that are deployed in the environment is extremely important and of course the performance of a machine learning model will depend on the quality of the data that we collect or generate. So, making this further connection across data network resources and the structure of machine learning models I think that will open doors for further improvements.

What future predictions can you make about the research field you are working on and the telecommunications sector?

Telecommunications are really at the basis of what is called digital revolution and besides keywords and slogans, I really believe that this is the truth.

Telecommunications has become something essential in our daily life and what we probably need to do is to push telecommunications forward in order to let them become more and more adaptive to the context where we live and to the user experience.

It is important to be able to adapt to the heterogeneity of the devices but also of human beings because each individual has its own needs, habits and preferences.

So, it is very important that we adapt to people and to the context in general and I think there is still quite a long way to achieve this goal although, recently the research that has been done really contributed to that objective.