Computer Networking & Future Networking Architectures
The members of the pole address multiple facets of networking architectures’ evolution. Their activities span content, information and data centric dynamic networking for distributed applications, services, systems and networks. We currently work on three main subjects.
First, future networking architectures tend to operate today as loosely coupled, distributed and cooperative systems, and to rely more on microservices. The evolution of interfacing between all layers in the networking architectures is becoming consequently a major concern. This is especially true for layers L4 to L0 that are also constantly evolving towards distributed and disaggregated systems composed and adapted with overall awareness of applications as well as bearer availability and physical medium conditions.
Second, we contribute actively to making the networking intelligent and capable of learning and adapting dynamically by introducing and embedding machine learning and artificial intelligence capabilities in the networking architectures. We are also active in the application of model based test generation strategies in SDN. This concerns testing SDN components in isolation, e.g., SDN switches / controllers, or their compositions.
Third, the forwarding and routing of flows and packets within the infrastructures can be based on content, information, naming or addressing, with routing and forwarding combined with efficient resource optimization and management. Programmability, configuration and control of distributed networks and networking functions in the context of the “cloud, SDN and NFV” trilogy is also central to this research activity with focus on dynamic services composition.