pr\u00e9sentent<\/p>\n\n\n\n
Autoris\u00e9 \u00e0 pr\u00e9senter ses travaux en vue de l\u2019obtention du Doctorat de l’Institut Polytechnique de Paris, pr\u00e9par\u00e9 \u00e0 T\u00e9l\u00e9com SudParis en :<\/p>\n\n\n\n
le JEUDI 12 D\u00e9CEMBRE 2024 \u00e0 10h00<\/p>\n\n\n\n
\u00e0<\/p>\n\n\n\n
Amphith\u00e9\u00e2tre 4
T\u00e9l\u00e9com SudParis, 19, Place Marguerite Perey, 91120 PALAISEAU<\/p>\n\n\n\n
Membres du jury :<\/strong><\/p>\n\n\n\n M. Fr\u00e9d\u00e9ric LEHMANN<\/strong>, Full professor, T\u00e9l\u00e9com SudParis, FRANCE – Directeur de these R\u00e9sum\u00e9 :<\/strong><\/p>\n\n\n\n L’acc\u00e8s multiple non-orthogonal grant-free (GF-NOMA) s’impose progressivement comme une partie int\u00e9grante de la couche physique des syst\u00e8mes d’acc\u00e8s radio du futur. En permettant d’acc\u00e9der \u00e0 une station de base sans allocation explicite de ressources temps\/fr\u00e9quence\/code, GF-NOMA permet non seulement d’am\u00e9liorer l’efficacit\u00e9 spectrale, mais \u00e9galement de rendre possible des communications ultra fiables \u00e0 faible latence (URLLC). De telles exigences permettront de r\u00e9pondre aux enjeux sp\u00e9cifiques d’applications sans fil telles que l’internet des objets, la r\u00e9alit\u00e9 virtuelle, les jeux vid\u00e9o en ligne, les communications entre machines, v\u00e9hicules, etc. Cependant, GF-NOMA introduit un nouveau d\u00e9fi inexistant dans les syst\u00e8mes de communication classiques, \u00e0 savoir la d\u00e9tection d’activit\u00e9 des utilisateurs : en plus de l’estimation du canal, de la d\u00e9tection et du d\u00e9codage des utilisateurs interf\u00e9rant, la station de base r\u00e9ceptrice doit \u00eatre en mesure de proc\u00e9der \u00e0 leur classification en deux cat\u00e9gories : ceux qui sont actifs et transmettent et ceux qui ne le sont pas. La massivit\u00e9 du syst\u00e8me, l’absence de contr\u00f4le de puissance \u00e0 l’\u00e9mission et\/ou d’orthogonalit\u00e9 des s\u00e9quences pilotes des utilisateurs sont autant de caract\u00e9ristiques qui compliquent les traitements en r\u00e9ception. Cette th\u00e8se a pour th\u00e8me g\u00e9n\u00e9ral l’\u00e9tude de nouvelles m\u00e9thodes statistiques bas\u00e9es sur des algorithmes \u00e0 passage de messages sur des graphes factoriels (factor graphs) appropri\u00e9s afin de traiter conjointement toutes ces t\u00e2ches au niveau du r\u00e9cepteur. Sont \u00e9tudi\u00e9es plus pr\u00e9cis\u00e9ment : – une m\u00e9thode (1) d’inf\u00e9rence bay\u00e9sienne hybride \u00e0 base de l’algorithme de propagation de croyance (belief propagation algorithm, BP) et de l’algorithme de propagation de l’esp\u00e9rance (expectation propagation algorithm, EP) pour r\u00e9soudre le probl\u00e8me conjoint de d\u00e9tection d’activit\u00e9, estimation de canal, et d\u00e9tection multi-utilisateur dans un syst\u00e8me GF-NOMA synchrone avec absence de contr\u00f4le de puissance \u00e0 l’\u00e9mission, s\u00e9quences pilotes orthogonales et antennes r\u00e9ceptrices multiples. En introduisant un crit\u00e8re d’approximation pour exprimer le passage de messages sous forme de lois gaussiennes, l’estimation du canal et la d\u00e9tection multi-utilisateurs peuvent \u00eatre trait\u00e9es efficacement par l’algorithme EP. Ceci s’av\u00e9rant impossible sous cette forme pour la d\u00e9tection d’activit\u00e9 des utilisateurs, un passage de messages sous forme BP est utilis\u00e9 \u00e0 cet effet. La m\u00e9thode propos\u00e9e inclut une \u00e9tape d’estimation des hyperparam\u00e8tres du mod\u00e8le que sont l’\u00e9nergie des signaux re\u00e7us et la corr\u00e9lation spatiale entre les antennes r\u00e9ceptrices. Une variante \u00e0 complexit\u00e9 r\u00e9duite ignorant la corr\u00e9lation spatiale entre antennes r\u00e9ceptrices est \u00e9galement propos\u00e9e ; – une m\u00e9thode (2) d’inf\u00e9rence bay\u00e9sienne \u00e0 base de l’algorithme EP exploitant des m\u00e9thodes d’analyse complexe (d\u00e9riv\u00e9es de Wirtinger) permettant de traiter la d\u00e9tection d’activit\u00e9 des utilisateurs \u00e9galement sous la forme d’un algorithme \u00e0 passage de messages gaussiens ; – une m\u00e9thode (3) faisant pr\u00e9c\u00e9der la m\u00e9thode (2) d’une m\u00e9thode d’acquisition comprim\u00e9e bay\u00e9sienne charg\u00e9e de l’estimation initiale du canal et de l’activit\u00e9 des utilisateurs dans le contexte complexifi\u00e9 d’un acc\u00e8s massif avec s\u00e9quences pilotes des utilisateurs non-orthogonales. L’\u00e9valuation par simulations de ces diff\u00e9rentes m\u00e9thodes est effectu\u00e9e dans le cas particulier d’un syst\u00e8me GF-NOMA synchrone par codage, entrelacement et modulation OFDM (GF-OFDM-IDMA). Les performances obtenues (mesur\u00e9es en termes de taux d’erreur binaire r\u00e9siduel pour la d\u00e9tection et le d\u00e9codage, d’erreur quadratique moyenne pour l’estimation de canal, et de probabilit\u00e9s de fausse alarme et de non-d\u00e9tection pour la d\u00e9tection d’activit\u00e9) se comparent favorablement par rapport \u00e0 celles obtenues avec des m\u00e9thodes classiques publi\u00e9es dans la litt\u00e9rature. Mots cl\u00e9s : NOMA, grant-free, acc\u00e8s massif, OFDM, graphes factoriels, algorithmes \u00e0 passage de messages, propagation de croyance, propagation de l’esp\u00e9rance. Grant-free non-orthogonal multiple access (GF-NOMA) is gradually becoming an integral part of the physical layer of future radio access systems. By allowing access to a base station without explicit allocation of time\/frequency\/code resources, GF-NOMA not only improves spectral efficiency, but also enables ultra-reliable low latency communications (URLLC) . Such requirements will make it possible to meet the specific challenges of wireless applications such as the Internet of Things, virtual reality, online video games, communications between machines, vehicles, etc. However, GF-NOMA introduces a new challenge that does not exist in conventional communication systems, namely user activity detection: in addition to channel estimation, detection and decoding of interfering users, the base station receiver must be able to classify them into two categories: those who are active and transmitting and those who are not. The massiveness of the system, the absence of power control on transmission and\/or orthogonality of user pilot sequences are all characteristics which complicate processing at the receiver. The general subject of this thesis is the study of new statistical methods based on message passing algorithms on appropriate factor graphs in order to jointly handle all these tasks at the receiver level. Are studied more precisely: – a method (1) of hybrid Bayesian inference based on the belief propagation algorithm (BP) and the expectation propagation algorithm (EP) to solve the problem of joint activity detection, channel estimation, and multi-user detection in a synchronous GF-NOMA system with no transmit power control, orthogonal pilot sequences and multiple receiver antennas. By introducing an approximation criterion to express message passing as Gaussian laws, channel estimation and multi-user detection can be efficiently processed by the EP algorithm. This proving impossible in this form for detecting user activity, message passing in BP form is used for this purpose. The proposed method includes a step of estimating the hyperparameters of the model, which are the energy of the received signals and the spatial correlation between the receiving antennas. A reduced complexity variant ignoring the spatial correlation between receiving antennas is also proposed; – a method (2) of Bayesian inference based on the EP algorithm exploiting complex analysis methods (Wirtinger derivatives) making it possible to process user activity detection also in the form of a Gaussian message passing algorithm; – a method (3) preceding method (2) with a Bayesian compressed acquisition method responsible for the initial estimation of the channel and user activity in the more complex context of massive access with non-orthogonal pilot sequences for the users. The evaluation by simulations of these different methods is carried out in the particular case of a synchronous GF-NOMA system by coding, interleaving and OFDM modulation (GF-OFDM-IDMA). The performance obtained (measured in terms of residual bit error rate for detection and decoding, root mean square error for channel estimation, and false alarm and missed-detection probabilities for activity detection) compare favorably with those obtained with traditional methods published in the literature. Keywords: NOMA, grant-free, massive access, OFDM, factor graphs, message passing algorithms, belief propagation, expectation propagation.<\/p>\n","protected":false},"excerpt":{"rendered":" L’Ecole doctorale : Ecole Doctorale de l’Institut Polytechnique de Paris et le Laboratoire de recherche SAMOVAR – Services r\u00e9partis, Architectures, Mod\u00e9lisation, Validation, Administration des R\u00e9seaux pr\u00e9sentent l\u2019AVIS DE SOUTENANCE de Monsieur Fakher SAGHEER Autoris\u00e9 \u00e0 pr\u00e9senter ses travaux en vue de l\u2019obtention du Doctorat de l’Institut Polytechnique de Paris, pr\u00e9par\u00e9 \u00e0 T\u00e9l\u00e9com SudParis en : […]<\/p>\n","protected":false},"author":4,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"ocean_post_layout":"","ocean_both_sidebars_style":"","ocean_both_sidebars_content_width":0,"ocean_both_sidebars_sidebars_width":0,"ocean_sidebar":"","ocean_second_sidebar":"","ocean_disable_margins":"enable","ocean_add_body_class":"","ocean_shortcode_before_top_bar":"","ocean_shortcode_after_top_bar":"","ocean_shortcode_before_header":"","ocean_shortcode_after_header":"","ocean_has_shortcode":"","ocean_shortcode_after_title":"","ocean_shortcode_before_footer_widgets":"","ocean_shortcode_after_footer_widgets":"","ocean_shortcode_before_footer_bottom":"","ocean_shortcode_after_footer_bottom":"","ocean_display_top_bar":"default","ocean_display_header":"default","ocean_header_style":"","ocean_center_header_left_menu":"","ocean_custom_header_template":"","ocean_custom_logo":0,"ocean_custom_retina_logo":0,"ocean_custom_logo_max_width":0,"ocean_custom_logo_tablet_max_width":0,"ocean_custom_logo_mobile_max_width":0,"ocean_custom_logo_max_height":0,"ocean_custom_logo_tablet_max_height":0,"ocean_custom_logo_mobile_max_height":0,"ocean_header_custom_menu":"","ocean_menu_typo_font_family":"","ocean_menu_typo_font_subset":"","ocean_menu_typo_font_size":0,"ocean_menu_typo_font_size_tablet":0,"ocean_menu_typo_font_size_mobile":0,"ocean_menu_typo_font_size_unit":"px","ocean_menu_typo_font_weight":"","ocean_menu_typo_font_weight_tablet":"","ocean_menu_typo_font_weight_mobile":"","ocean_menu_typo_transform":"","ocean_menu_typo_transform_tablet":"","ocean_menu_typo_transform_mobile":"","ocean_menu_typo_line_height":0,"ocean_menu_typo_line_height_tablet":0,"ocean_menu_typo_line_height_mobile":0,"ocean_menu_typo_line_height_unit":"","ocean_menu_typo_spacing":0,"ocean_menu_typo_spacing_tablet":0,"ocean_menu_typo_spacing_mobile":0,"ocean_menu_typo_spacing_unit":"","ocean_menu_link_color":"","ocean_menu_link_color_hover":"","ocean_menu_link_color_active":"","ocean_menu_link_background":"","ocean_menu_link_hover_background":"","ocean_menu_link_active_background":"","ocean_menu_social_links_bg":"","ocean_menu_social_hover_links_bg":"","ocean_menu_social_links_color":"","ocean_menu_social_hover_links_color":"","ocean_disable_title":"default","ocean_disable_heading":"default","ocean_post_title":"","ocean_post_subheading":"","ocean_post_title_style":"","ocean_post_title_background_color":"","ocean_post_title_background":0,"ocean_post_title_bg_image_position":"","ocean_post_title_bg_image_attachment":"","ocean_post_title_bg_image_repeat":"","ocean_post_title_bg_image_size":"","ocean_post_title_height":0,"ocean_post_title_bg_overlay":0.5,"ocean_post_title_bg_overlay_color":"","ocean_disable_breadcrumbs":"default","ocean_breadcrumbs_color":"","ocean_breadcrumbs_separator_color":"","ocean_breadcrumbs_links_color":"","ocean_breadcrumbs_links_hover_color":"","ocean_display_footer_widgets":"default","ocean_display_footer_bottom":"default","ocean_custom_footer_template":"","ocean_post_oembed":"","ocean_post_self_hosted_media":"","ocean_post_video_embed":"","ocean_link_format":"","ocean_link_format_target":"self","ocean_quote_format":"","ocean_quote_format_link":"post","ocean_gallery_link_images":"on","ocean_gallery_id":[],"footnotes":""},"categories":[286,543],"tags":[],"class_list":["post-6722","post","type-post","status-publish","format-standard","hentry","category-fractualites-ennews-fr","category-seminaire-istec","entry"],"_links":{"self":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/6722","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/comments?post=6722"}],"version-history":[{"count":1,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/6722\/revisions"}],"predecessor-version":[{"id":6723,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/6722\/revisions\/6723"}],"wp:attachment":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/media?parent=6722"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/categories?post=6722"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/tags?post=6722"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
M. Jean-Marie GORCE<\/strong>, Professeure des universit\u00e9s, INSA Lyon, FRANCE – Rapporteur
M. Jean-Pierre CANCES<\/strong>, Professeur des universit\u00e9s, Universit\u00e9 de Limoges, FRANCE – Rapporteur
M. Didier LE RUYET<\/strong>, Professeure des universit\u00e9s, CNAM Paris, FRANCE – Examinateur
Mme Lina MROUEH<\/strong>, Professeure, ISEP, FRANCE – Examinateur
M. Antoine BERTHET<\/strong>, Full professor, Centrale Supelec, FRANCE – CoDirecteur de these<\/p>\n\n\n\n\u00ab M\u00e9thodes statistiques bay\u00e9siennes pour la d\u00e9tection conjointe des activit\u00e9s des utilisateurs, l\u2019estimation des canaux et le d\u00e9codage des donn\u00e9es dans les r\u00e9seaux sans fil dynamiques \u00bb<\/h2>\n\n\n\n
pr\u00e9sent\u00e9 par Monsieur Fakher SAGHEER<\/h2>\n\n\n\n
Abstract :<\/strong><\/p>\n\n\n\n