PhD «State evaluation, reconfiguration and resynchronization of Network Digital Twins » M/F

about the role

Your role is to work on the PhD thesis entitled “State evaluation, reconfiguration and resynchronization of Network Digital Twins”

Global Context and Problematic of the Topic

Digital twins (DTs) for networks are part of the major strategic technological trends at Orange.

For Orange, the continuous evolution of networks leads to an increasing complexity in managing physical entities (optical/mobile core network), their physical environment (decentralized, shared), and software. This complexity results in difficulties in automating networks, both in nominal operation (monitoring) and in incident management.

To address this evolution, it is necessary to have a holistic view of networks. Network DTs are seen as the central enabler to provide this vision, thus moving towards the capabilities of next-generation autonomous networks: monitoring, simulation, reconfiguration[1], orchestration of heterogeneous networks, impact analysis on different network layers[2], anomaly detection[3]...

Scientific objective: results and challenges to overcome

This work will be conducted in the technological environment of the Thing’in digital twin platform developed by Orange, and the open-source Telco Cloud platform Sylva, of which Orange is one of the founders and contributors (https://sylvaproject.org/about-sylva/).

The main ambition is, through the DT, to detect drifts between a target infrastructure (Sylva) and the actually deployed infrastructure, and to propose reconfigurations to resynchronize the two elements (alignment of IT and field).

The major objectives are to define and develop:

Data models that will represent, within the DTs, the states of networks, network functions, target infrastructures (the IT) and those actually deployed (the field).
Continuous evaluation approaches, including predictive/simulated, for state analysis and drift detection at the DT level.
Resynchronization, taking into account technical constraints (heterogeneity of networks, access constraints, latency, resources, deployment, scalability...).

Methodological approach

State of the art of DTs, autonomous networks, virtualized networks.

Development of architectures, data models, algorithms, in collaboration with research and business experts. The work may build on previous studies: historization of DTs, complex DTs, predictive approaches for DT graphs.
Implementation through Thing’in and Sylva of DT prototypes in a use case from an internal Orange project, in a laboratory or at a large scale (testbed GRID5000/SLICES-FR).

The work should lead to publications, and possibly patents, in the fields of network management, digital twins, AI, distributed systems, and cloud computing.