Job Description

Resource allocation and signaling problems in 6G wireless communication networks

Funding project: “6th Generation Wireless Communications Networks: New Concepts, Algorithms and Applications” (6th. Geração: Novos Conceitos, Algoritmos e Aplicações), FAPESP / MCTI / CGI - 2022
Supervisors: Prof. Rodrigo de Lamare (PUC-Rio, Univ. of York), Prof. Lukas Landau (PUC-Rio) and Prof. E. Veronica Belmega (Gustave Eiffel Univ.)
Home institution: Pontifícal Catholic University of Rio de Janeiro (PUC-Rio), Brazil
Host institution: Gustave Eiffel Univ., Greater Paris, France
The candidate will be based at PUC-Rio, Rio de Janeiro, Brazil and will also have the opportunity to spend part of the fellowship (according to FAPESP’s rules with a fellowship in Euro) at the Université Gustave Eiffel and its graduate engineering school ESIEE Paris in the Greater Paris area in France.
Contract duration and salary: up to 4 years with a fellowship of approximately BRL 9.1 per month (tax free) and funds for travel and research expenses. The start date can be as early as January 2024. During the visits in France, the fellowship will be converted to Euro.
Pre-requisites: Candidates must hold a PhD degree in electrical engineering (with a focus on telecommunications and/or signal processing) or applied mathematics. Having a strong mathematical background and/or computer literacy skills (Python, C++, MatLab, etc.) is a definite plus.
How to apply: Candidates should contact the supervisors via email (rodrigo.delamare@york.ac.uk, landau@puc-rio.br, veronica.belmega@esiee.fr) providing a short motivation and a detailed academic-style CV including the list of publications. Applications will be received until the position is suitably filled.

Research summary
6G communications networks will have challenging requirements in terms of resource allocation such as power, frequency bands, users, access point channels and repeaters, and of signaling information containing key parameters used in the coordination of network elements such as devices, cells and clouds. In this context, clouds (CRANs and FRANs) have the potential to mitigate interference between access points through coordinated resource allocation between devices and access points.
The vast majority of existing resource allocation policies make simplifying assumptions regarding the temporal variability of the network and the channel state information knowledge available at the transmitter: static networks and perfect channel information or stochastic networks and perfect channel distribution information. However, in future IoT and 6G networks these assumptions may not necessarily be met, due to the network density, their heterogeneity, device mobility and connectivity patterns, which could lead to an unpredictable and non-stationary underlying dynamics. New tools that go beyond classical and stochastic optimization and game theory are needed to solve such problems.
Objectives
The selected candidate will develop innovative and efficient resource optimization algorithms suitable to this context with a cost-effective use of network signaling. The main idea is to exploit novel optimization approaches and robust techniques to deal with channel and other parameter uncertainties and temporal variability coupled with innovative approaches to reduce the signaling load in networks. In addition, the researcher is expected to develop mathematical models and tools to rigorously analyze the performance of such algorithms. Also, extensive numerical simulations will have to be performed to evaluate them.
The scientific results will be published in the form of articles in top-tier international conference proceedings (IEEE Globecom, ICC, SPAWC, etc.) and international journals (IEEE Trans. on Wireless Commun., Trans. on Signal Processing, etc.).
At last, the postdoctoral researcher will be trained to apply for professorships and to endorse the position of professor, advisor and research supervisor.
Workplan
During the first part (two years) of the project, the postdoctoral researcher will investigate resource allocation problems (energy, access points, channels, frequencies, etc.) that can be solved using efficient algorithms with reduced use of signaling information. A first idea is to develop heuristic algorithms and greedy-type techniques to solve such problems with low computational complexity and performing close to the optimal solution [1]. In addition, the researcher will investigate robust resource allocation techniques in the presence of uncertainty and reduced signaling in networks. In this case, robust worst-case optimization techniques and more recent approaches that exploit a priori knowledge about the system parameters will be considered [2], [3]. The reduced signaling load in the network corresponds to the need to exchange information about communication channels and various important parameters between devices, access points and clouds. The project proposes an investigation into ways of reducing the signaling load on the network, including the use of access points with limited access to clouds, the compression of information, and the use of learning algorithms that use binary signaling or incomplete information about the parameters.
During the second part (two years) of the project, the postdoctoral researcher will explore the online optimization toolbox [4,5] going beyond classical and stochastic optimization, allowing for variations in the problem that are completely arbitrary, accounting for exogenous parameters that may be unpredictable and non-stationary. This is of particular relevance for future IoT and 6G networks, which are dense and time-varying networks, and in which the heterogeneity of the devices serving various applications, their mobility and connectivity patterns can lead to an arbitrarily time-varying wireless environment [6,7]. The goal in such settings is to design low-complexity and energy-efficient algorithms for future wireless networks (IoT, 6G etc.) exploiting online optimization and reinforcement learning to deal with such system dynamics and based on limited and reduced feedback information (down to a single ACK/NACK bit).
Throughout the post-doctoral period, the researcher will be expected to help lead the implementation of the algorithms in software-defined radio platforms, helping and guiding undergraduate students working on the platform.
References
[1] S. Mashdour, R. C. de Lamare and J. P. S. H. Lima, "Enhanced Subset Greedy Multiuser Scheduling in Clustered Cell-Free Massive MIMO Systems," in IEEE Communications Letters, vol. 27, no. 2, pp. 610-614, Feb. 2023.

[2] H. Ruan and R. C. de Lamare, "Robust Adaptive Beamforming Based on Low-Rank and Cross-Correlation Techniques," in IEEE Transactions on Signal Processing, vol. 64, no. 15, pp. 3919-3932, 1 Aug.1, 2016.

[3] H. Ruan and R. C. de Lamare, "Distributed Robust Beamforming Based on Low-Rank and Cross-Correlation Techniques: Design and Analysis," in IEEE Transactions on Signal Processing, vol. 67, no. 24, pp. 6411-6423, 15 Dec.15, 2019.

[4] S. Shalev-Shwartz (2011). “Online learning and online convex optimization”. Foundations and trends in Machine Learning, 4(2), 107-194.

[5] E.V. Belmega, P. Mertikopoulos, and R. Negrel, "Online convex optimization in wireless networks and beyond: The feedback - performance trade-off", invited paper at RAWNET intl. workshop in conjunction with WiOpt, Turin, Italy, Sep. 2022

[6] O. Bilenne, P. Mertikopoulos, and E.V. Belmega, "Fast Optimization with Zeroth-order Feedback in Distributed Multi-User MIMO Systems", IEEE Trans. on Signal Processing, vol. 68, pp.6085-6100, Oct. 2020.

[7] A. Marcastel, E.V. Belmega, P. Mertikopoulos, and I. Fijalkow, "Online power optimization in feedback-limited, dynamic and unpredictable IoT networks", IEEE Trans. on Signal Processing, vol. 67, no. 11, pp. 2987-3000, Jun. 2019.

Follow @eurasip