Pause
Lecture
Moteur de recherche d'offres d'emploi CEA

D-band beam alignment algorithms H/F


Détail de l'offre

Informations générales

Entité de rattachement

Le Commissariat à l'énergie atomique et aux énergies alternatives (CEA) est un organisme public de recherche.

Acteur majeur de la recherche, du développement et de l'innovation, le CEA intervient dans le cadre de ses quatre missions :
. la défense et la sécurité
. l'énergie nucléaire (fission et fusion)
. la recherche technologique pour l'industrie
. la recherche fondamentale (sciences de la matière et sciences de la vie).

Avec ses 16000 salariés -techniciens, ingénieurs, chercheurs, et personnel en soutien à la recherche- le CEA participe à de nombreux projets de collaboration aux côtés de ses partenaires académiques et industriels.  

Référence

2021-18410  

Description du poste

Domaine

Composants et équipements électroniques

Contrat

Stage

Intitulé de l'offre

D-band beam alignment algorithms H/F

Sujet de stage

D-band communications impose the use of very directive antennas. As the antennas may move due to wind for example, the beam to use must be changed in real-time so that the antennas are always aligned. The student will have first to imagine one or several algorithms that allow to anticipate the antenna beam to use, based on previous used beams. The student will then code one of these algorithms (MATLAB or C). A solution that can be considered is a machine learning type algorithm. With this kind of algorithm, the antenna could learn from the past events to predict the future ones. The algorithm is intended to run on a hardware board. No particular skills in board design or hardware language is required because the algorithm will be implemented on a RISC-V softcore implemented on an FPGA. This type of implementation allows the algorithm to be written in MATLAB or C language.

Durée du contrat (en mois)

6

Description de l'offre

The DRAGON European project aims at demonstrating a high capacity D-band (130-174.8 GHz) wireless backhaul solution able to address the needs of 5G transport networks. A pole-mounted transceiver will be designed and implemented, including 1024-element phased array antennas. The target distance between the pole-mounted transceiver and the gNB – up to 1 km – imposes very high gain antennas, and therefore the use of very narrow beams (in the order of 3.5° beamwidth). This feature may cause misalignment between transmit and receive beams when poles vibrate. In this context, a dedicated module, the Phased Array Manager (PAM) will be responsible to control the antenna drivers and manage the antenna beam direction at all time to ensure a good enough beam alignment.

In DRAGON, CEA is responsible for the PAM design, implementation and integration into the transceiver. The strategy to determine the best beam is to use a dedicated time slot in the frame – the Beam Alignment (BA) field – during which the received power with one selected beam is compared to the received power with the current beam. The beam that provides the highest power between both measures is then chosen for transmission in the next frames. The number of beams from which the PAM can choose the beam to be assessed during the BA field is limited. This set is called the codebook.

As only one beam can be tested during a BA field, a particular attention must be paid in the choice of this beam in the codebook. A bad strategy can lead to link outage during several frames. When the frequency and the amplitude of the vibrations of the pole-mounted antenna are slow, it was shown that a very simple, systematic, strategy can be used. Nevertheless, if the frequency and/or the amplitude of the vibrations increase, the direction to test must be anticipated.

This is the objective of the internship. The student will have first to imagine one or several algorithms that take into account the past selected beams in order to anticipate the next beam to assess. The student will then code one of these algorithms (MATLAB or C). A solution that can be considered is a machine learning type algorithm. With this kind of algorithm, the PAM could learn from the past events to predict the future ones. The algorithm is intended to run a hardware board. No particular skills in board design or hardware language is required because the algorithm will be implemented on a RISC-V softcore implemented on an FPGA. This type of implementation allows the algorithm to be written in MATLAB or C language.

For these tasks, the student will be able to rely on experts in the LSHD laboratory in each field of study (algorithm design,machine learning and implementation on board). If the results are satisfactory, the algorithm will be implemented in the final demonstrator of the project and will lead to the writing and submission of a paper.

Moyens / Méthodes / Logiciels

MATLAB, C.

Profil du candidat

Required skills:

· Algorithmics

· Languages: Matlab and/or C

· English (reading and writing)

Telecom would be a plus.

Localisation du poste

Site

Grenoble

Localisation du poste

France, Auvergne-Rhône-Alpes, Isère (38)

Ville

Grenoble

Critères candidat

Diplôme préparé

Bac+5 - Diplôme École d'ingénieurs

Possibilité de poursuite en thèse

Oui

Demandeur

Disponibilité du poste

01/02/2022