Stage AI for QoS Prediction and Network Reconfiguration in Industrial Environment H/F
Detail de l'annonce :
The 5G communication standard is the prerequisite of the factory of
the future. By enabling high reliability, low latency and security
features, 5G networks introduce more flexibility and production in
manufacturing tasks and support wireless connectivity in industrial
environment. In this context of critical Internet of Things (IoT) as
industrial automation use cases, the prediction of Quality of Service
(QoS) allows the notification in advance of a potential change in the
QoS, thus avoiding reduction of undesired behaviour and damages.
In case a degraded QoS is predicted, different countermeasures may be
applied at the application level such as slowing down the motion or
process speed, update the robots trajectories. Another option is to
reconfigure the network parameters as the resource reservation. In the
proposed internship, we focus on the second option: the network
reconfiguration.
During the internship, the main goals to achieve are:
1. To enhance developed AI techniques based on neural networks and
clustering (and to investigate other algorithms) to predict the
degradation of quality of service, in particular the transmission
latency and packets loss.
1. To propose the adequate network optimization based on QoS
predictions and traffic information in order to achieve the QoS
targets. A reinforcement learning approach may be followed.
ABOUT YOU
An internal system level simulator developed in C++ for 5G Ultra
Reliable Low Latency Communication (URLLC) will be used during the
internship for data generation.
QoS prediction techniques (Neural Networks, Clustering, Regression)
developed in Python in previous works can be reused in the first steps
of the internship.
ADDITIONAL INFORMATION
Vous serez un acteur direct des transformations technologiques en
cours qui accompagne les changements sociétaux (inclusion numérique,
santé connectée, villes intelligentes, …).
Vous intégrerez une équipe dynamique et multi-culturelle avec des
experts radio internationalement reconnus.
Vous développez une expertise pointue sur les normes, sur les
équipements mobiles et réseaux d’accès mobile (RAN), sur les
enjeux d’interopérabilités des systèmes radio, ainsi que sur la
stratégie de déploiement des différentes filiales, avec un focus
particulier sur l’analyse des performances radio mobile 5G (MiMo, IA
pour le RAN, 6G,…).
DEPARTMENT
Au sein du département RIO (Radio Innovation), l’équipe RAP (RAdio
Performance) est en charge de la conception et de l’évaluation de
performances des solutions technologiques radio-mobile de demain. Ses
études se portent aujourd’hui principalement sur l’évaluation
des performances des réseaux mobiles 5G, sur la stratégie spectre et
sur la conception de la future technologie 6G, dont l’apport de
l’intelligence artificielle constitue une brique essentielle.
CONTRACT
Internship
Level : Bac+5