Traffic Forecasting for Resource Allocation in B5G/6G Networks

ETIS is a joint research department between CYU Cergy Paris University, ENSEA Graduate School of Electrical Engineering and CNRS/INS2I.

ETIS develops research in the field of the theory of information with both theoretical and experimental activities in order to allow information processing systems to acquire capacities of autonomy. Autonomy is considered both in terms of learning and adaptation to the environment (including users) as well as making decision that includes low energy consumption and computing power for example.

ETIS designed systems perform intelligent processing which is adaptable to increasing complexity. The concerned areas are reconfigurable chip systems, data analysis, image indexing, developmental robotics, information theory and telecommunications. Learning and adaptation algorithms based on data constitue the core of the developed systems.

The ETIS laboratory is at the heart of the current AI revolution Learning and adaptation algorithms based on data constitue the core of the developed systems.

The next generation networks have been defined to intentionally meet a wide range of various vertical industries’ requirements. They promise to provide anytime and anywhere connectivity for everything from personal smart mobiles and devices to machines and sensors, offer limitless experience in mobile broadband, and give the ability to support critical machine-to-machine communications with low latency and ultra-high reliability. Indeed, next generation networks will support multiple services with specific performance requirements in highly heterogeneous environments. Therefore, it is important for the wireless service providers to allocate and utilize the network resources efficiently. To this end, resource allocation management in wireless communication systems will require accurate traffic analysis and prediction to guarantee an optimal resource utilization. To efficiently fulfill the requirements of different applications, some works in the literature developed algorithms for traffic prediction based on machine learning.

This internship will provide innovative resource allocation frameworks that allow network operators to allocate resources adaptively according to the time-varying requirements, the services’ demand, and the available resources.

• The candidate should be enrolled in a master (or bac + 5) in wireless communication, machine learning, or any related field.
• Good knowledge of Python and/or Matlab.
• Knowledge of machine learning and optimization tools.