Open Positions

Call for post-doctoral applications (PDF version)

The SARA team (http://www.laas.fr/SARA-EN/) at LAAS-CNRS (http://www.laas.fr/), Toulouse, France, is seeking a 1-year postdoc (possible extension to 2 years) to work on the research project ONSET. The project is funded by the French Defense Agency and is to be started in November 2020. The main objective of ONSET is to develop new optimization methods for the optimal design of network slices as well as new data-driven learning mechanisms for their dynamic reconfiguration (auto-scaling, load balancing and migration). The position is to be filled by February 2021 at latest. The selected candidate will join a research group composed of Olivier Brun (http://homepages.laas.fr/brun/), Pascal Berthou (http://homepages.laas.fr/berthou/) and Slim Abdellatif.  The monthly gross salary is around  2,500 €. The selected candidate will be encouraged to apply to CNRS and Université de Toulouse in order to obtain a permanent appointment at LAAS.

Title: Optimal Network Slice managEmenT

Context: Whilst earlier network generations have been designed as general-purpose connectivity platforms, future networks will have to support the specific needs of a plethora of new services developed by a variety of vertical industries (e.g., automotive, energy, food and agriculture, healthcare, or public transportation). These services will have a diverse range of requirements ranging from high reliability to ultra-low latency going through high bandwidth and mobility. 

Network slicing has recently emerged as a key technique for sharing a single physical network infrastructure between multiple logical and independent networks that are configured to effectively meet the various services requirements. These logical networks, which are called network slices, can be used by third parties that do not possess their own network infrastructure to provide network services tailored to their specific needs. Each network slice is implemented as a chain of virtualized network functions (VNF) running on virtual machines inside an operator network or in centralized clouds. Each logical network is isolated from the others, independently managed/controlled, and can adapt quickly to the demand.

Orchestration is a key process for network slicing, which involves optimally selecting the physical resources to be used to fully meet service requirements (SLA: Service Level Agreement). Orchestration also allows performance to be maintained when demand evolves by monitoring performances and by reconfiguring as needed. The orchestration of logical networks however raises a number of conceptual difficulties. The objective of the ONSET project is to address these difficulties by developing:

  • New methods for planning the capacity of network slices as well as for selecting the physical resources on which they are deployed, considering quality of service and survivability requirements as well as priority between network slices,
  • Data-driven learning mechanisms for the dynamic reconfiguration of network slices (auto-scaling, load balancing and migration) so as to maintain satisfactory performance in spite of extremely variable load conditions,
  • A software environment for the simulation, performance evaluation and optimization of virtualized networks over a multi-domain and multi-technology physical infrastructure. This environment will integrate all the algorithms developed during the project for the planning and dynamic reconfiguration of network slices.

 

Candidate Profile

The candidate must have:

- a PhD degree in computer science or a related field, 

- excellent knowledge in the domain of communication networks, 

- good skills in mathematical optimization theory, machine learning or stochastic modelling (queueing theory, MDPs,...),

- strong computer programming skills,

- good organizational and communication skills.  

 

Application

Each application must contain the following documents: 

- a detailed CV 

- a cover letter with a short description of previous works 

- one or two references with contact addresses 

 

Applications should be sent electronically to Olivier Brun (brun@laas.fr).