- Network monitoring, performance evaluation, network management, quality of service, quality of experience anomaly detection, network security, ressource allocation
- Bayesian inference, MCMC methods, (hidden) Markov chains, Markov decision processes, reinforcement learning, optimisation, game theory, distributed computing, stochastic approximations
- Network virtualization, Distributed Ledger Technologies, Internet 5G
The Maths&Net team aims to design, describe, manage, secure and control various aspects of networks, in particular telecommunications networks. The team also works on other types of networks such as distributed ledgers or social networks.
In our work on telecommunication networks, the objective is to contribute to the quality of service at network level, to the quality of experience of applications, to the availability of network resources (or of resources accessible through the network) and to network security (detection of cyber attacks).
Our work is at the interface between the field of networks and distributed systems (from an application point of view), and the field of machine learning and operational research (from a theoretical point of view).
Given the abundance but also the specificity of network data, the potential offered by advanced data analysis techniques, and the flexibility introduced into network management by virtualization techniques, our objective is to make the most of data analysis to design high-performance, secure and high-availability networks.