Post-doctoral fellow in Access and management of non-persistent data in the cloud/edge at Télécom Paris - CDD 24 months

Job description

Télécom Paris, a school of the IMT (Institut Mines-Télécom) and a founding member of the Institut Polytechnique de Paris, is one of the top 5 French general engineering schools. An institution on a "human scale" but with a strong international component, Télécom Paris is recognized for its proximity to companies. This public school guarantees excellent employability in all sectors and is the leading engineering school for the entire digital vertical (from hardware layers to uses).

With its excellent teaching and innovative pedagogy, Télécom Paris is at the heart of a unique innovation ecosystem, based on interaction and the importance of project mode in its training on the one hand, and its interdisciplinary research on the other. Its teacher-researchers are affiliated with two research laboratories: on the one hand, the LTCI laboratory, which is presented by the HCERES as a flagship unit in the field of digital sciences with remarkable international influence; and on the other hand, the i3 laboratory, Interdisciplinary Institute of Innovation (I3 - UMR 9217 of the CNRS), which pursues a multidisciplinary research program focused on innovation in the framework of a collaboration with the École Polytechnique and Mines ParisTech

Based in Palaiseau, at the heart of the Institut Polytechnique campus alongside the École Polytechnique, ENSTA, Télécom Sud Paris and ENSAE, Télécom Paris also has a Paris-based incubator at the heart of the French start-up ecosystem.

Télécom Paris is positioned as an open-air laboratory for all the major technological and societal challenges: artificial intelligence, quantum computing, IOT, cybersecurity, large-scale digital equipment (Cloud), 5G/6G, Green IT.


In modern and post-5G networks, the massive connectivity of heterogeneous devices (from 5G base stations to tiny IoT devices) can generate a large amount of data that can be stored within, or transferred across the network. In highly distributed architectures such as the Mobile Edge Computing (MEC), data can be ubiquitously accessed from any device in the network, which raise the questions about how to access and manage the data. For example, a data stream (e.g., measurements from an IoT sensor, or a file sent towards a remote server) can be locally generated at the end device, and then reach an edge or cloud server, according to the constraints of the required processing. A recently adopted paradigm (called Network Function Virtualization, or NFV) requires the data stream to further traverse one or more virtual network functions (VNFs) to be processed at the edge or cloud server. At this stage, the data stream can be stored persistently, either in the cloud, or within the edge equipment close to the actual source. The data can be then accessed from anywhere in the network, independently from the actual location where the data will be persisted. Alternatively, the different data streams to be processed in each VNF can leave some “breadcrumbs” that can be deleted after the end-to-end exchange is finalized. In other words, this soft state is non-persistently stored inside one (or more) VNF(s). A soft-state approach is sometimes preferable w.r.t. the common persistent storage. For instance, it can be useful to perform some pre-processing at the edge before sending the data to the cloud, in order to reduce the amount of data exchange (which lowers the network bandwidth usage) or storage occupancy (which limits the amount of storage required and reduces both costs and carbon footprint). Depending on the application, it can even be possible to keep the data streams from leaving the edge components in the general case, and perform an exchange towards the cloud only if a fine-grained analysis is needed (as an example, we can consider the anomaly detection problem, or the reaction following an alarm signal). Finally, the temporarily stored soft state can also be used to carry some metadata about the connection quality, additional measurements about the traffic rate or the network bandwidth and the utilization of the CPU (or other resources) of the devices traversed by the data stream.

Ephemeral data

This can be modeled as a problem of access and management of ephemeral data, in the sense that some processing can occur with non-persistent data that exist only in the context of an end-to-end interaction and can be afterwards safely removed. Our goal is to provide an architecture for soft-state access and management that can be used to quickly access the current state of the network elements, react to faults or find the optimal resource allocation. This approach comes with some challenges. First, a continuous fine-grained monitoring of the network state can be required. Furthermore, in the case of network faults of an entity (usually a VM or a container), some data can be lost or corrupted. Finally, such soft state can still leak some sensitive information that can be used by malicious external entities, and as such it is important to provide privacy and integrity.


1. The objective of this post-doc is to design and evaluate an architecture for the generation,

access and management of soft-state information in NFV systems. The architecture should

have the following features:

  • provide the interface for generating the requests about the soft state at the level of the VNF or the data stream itself.
  • provide the mechanisms to manage the soft-state: e.g., understand when the ephemeral data can be safely removed

2. As a driving use case, we consider the deployment of some algorithms for the fair resource sharing of the network resources that use the ephemeral data via the provided interface to continuously get some information from the network.

3. A proof of concept should be deployed at the end of the collaboration, with a potential

integration in an experimental platform for further experimental and performance analysis.


1.Carry out research assignments in the field of high-speed network architectures

Carry out research missions in the framework of the Department's contracts in the virtualised networks.

Participates in and ensures the production of project deliverables.

2.To ensure supervision and tutoring missions

Supervises student projects and engineering internships

3.Contribute to the reputation of the School, the Institut Mines-Télécom and the Institut Polytechnique de Paris

Publication of the results of its research work

Gives presentations and conferences

Involved in learned or professional societies

Maintain close relations with academic institutions, research centers and companies

4.Other activities

Participates in the scientific, pedagogical and management activities of the Department or Institution

Where appropriate, directs and manages the staff placed under his/her responsibility or supervision

Reports on the activities and results of the missions for which he/she is responsible

Job requirements



Essential Skills, Knowledge and Experience:

  • In-depth theoretical and practical knowledge in the field of network virtualisation and software and software network architectures (NFV, containerisation, placement)Implementation of qualitative studies
  • Fluency in English

Desirable skills, knowledge and experience:

  • Experience with broadband deployment tools (DPDK, netmap, MoonGen)

Skills and Abilities:

  • Ability to work in a team, dialogue and writing skills
  • interpersonal and pedagogical skills
  • Ability to summarise


Level of education and/or experience required:

Ph.D. or equivalent in the following fields (Computer Science, or in Computer engineering Computer engineering)

Contract type : CDD Post doctoral
Contract duration: 24 months or more

Location: Télécom Paris, 19 Place Marguerite Perey, Palaiseau 91120


The application include:

  • a detailed CV
  • a letter of motivation
  • any element considered useful for the examination of the application

Deadline for application August 12, 2022

Scientific contact : Leonardo LINGUAGLOSSA >

Administrative contact : Céline LAOUENAN >