PhD position (CIFRE) offer with Thales SIX GTS France on Artificial Intelligence for resource allocation in mobile clustered ad hoc networks.
Thales SIX GTS France is offering a CIFRE PhD position in the Waveform Design Department to research on Machine Learning techniques applied to resource allocation in mobile ad hoc networks. The topic focuses on the joint optimization of Physical layer and layer Data Link layer for resource allocation of the radio access. We consider advanced radio access with multi-user, multiple access capabilities such as NOMA (non-orthogonal multiple access) along with several MCS (modulation and coding schemes) including Hybrid ARQ (HARQ) and power adaptation. The optimization is multi-criteria under constraints taking into account the different traffic Quality of Service (QoS), finite length queues, propagation channel… It also needs to adapt over time to take into account the changing environment due to mobility.
Recent works (e.g. ) have shown that mathematical derivations using tricks of convex optimization allow to exactly solving subpart of the general problem, the solutions remain complex and not applicable to all cases. Thus, the goal of this thesis is to explore the application of Machine Learning techniques to develop solutions in order to solve the general problem. The use of these tools belongs to Artificial Intelligence (AI) which is a very hot topic in wireless communications today. Although not restrictive, it is already envisioned to model the problem using the Markov Decision Process (MDP) framework and to solve it by applying Deep Q (or reinforcement) learning techniques using approximations of the cost-value function by Neural Networks . Both offline and online training will be considered. Extension to time-varying number of users might be also considered if time allows.
 X. Leturc et al., “Energy-Efficient Resource Allocation for HARQ With Statistical CSI”, IEEE Tr Vehicular Technology, Dec. 2018.
 Z. Xiong et al., “Deep Reinforcement Learning for mobile 5G and beyond: fundamentals, applications, and challenges, IEEE Vehicular Technology Mag, Jun 2019.
Strong motivation and commitment, strong mathematical background, background in wireless communications preferred, good programming skills in Matlab and Python required.
Contract and salary
Three year duration contract with Thales SIX GTS France, net salary: around 1800€ per month.
Industry: Christophe le Martret (Thales Expert, HDR), Thales SIX GTS France, F-92622 Gennevilliers.
Laboratory: Philippe Ciblat (Professor), Télécom ParisTech & Institut Polytechnique de Paris, F-75013 Paris (F-91477 Palaiseau from Nov. 2019).
Send a résumé and cover letter to Christophe Le Martret: email@example.com
(c) GdR 720 ISIS - CNRS - 2011-2019.