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 VENDREDI 12 D\u00e9CEMBRE 2025 \u00e0 13h45<\/p>\n\n\n\n
\u00e0<\/p>\n\n\n\n
Amphith\u00e9\u00e2tre 3
T\u00e9l\u00e9com SudParis 19 place Marguerite Perey 91120 PALAISEAU<\/p>\n\n\n\n
https:\/\/telecom-paris.zoom.us\/j\/93461003734?pwd=yDpmnnaYViTFtUdJUcalm7LUCBcyeU.1<\/a> Membres du jury :<\/strong><\/p>\n\n\n\n M. Djamal ZEGHLACHE<\/strong>, Professeur, T\u00e9l\u00e9com SudParis, FRANCE – Directeur de th\u00e8se R\u00e9sum\u00e9 :<\/strong><\/p>\n\n\n\n La croissance exponentielle du trafic de donn\u00e9es et la complexit\u00e9 croissante des r\u00e9seaux de transport optique ont cr\u00e9\u00e9 un besoin imp\u00e9rieux d’architectures de contr\u00f4le de r\u00e9seau plus flexibles, automatis\u00e9es et intelligentes. Les syst\u00e8mes optiques traditionnels, monolithiques et propri\u00e9taires, ne peuvent plus r\u00e9pondre aux exigences de d\u00e9ploiement rapide de services, de mise \u00e0 l’\u00e9chelle et de gestion de r\u00e9seau autonome. Cette th\u00e8se aborde ces d\u00e9fis en d\u00e9veloppant un cadre complet pour le contr\u00f4le de l’infrastructure et l’orchestration de services de bout en bout dans les futurs r\u00e9seaux de transport. Ce contr\u00f4le et cette gestion sont accomplis par la convergence de la mise en r\u00e9seau d\u00e9finie par Software-Defined Networking (SDN), de la d\u00e9sagr\u00e9gation du r\u00e9seau et de l’intelligence fond\u00e9e sur l’apprentissage automatique (Machine Learning). La recherche pr\u00e9sente trois contributions principales qui, collectivement, permettent un fonctionnement autonome des r\u00e9seaux optiques. Premi\u00e8rement, une plateforme de contr\u00f4le pour le transport optique, ouverte, d\u00e9sagr\u00e9g\u00e9e et native pour le cloud est d\u00e9velopp\u00e9e, comprenant un contr\u00f4leur SDN de transport (T-SDNC) bas\u00e9 sur OpenDaylight et dot\u00e9 d’une architecture en microservices. Une am\u00e9lioration essentielle de Transport API (TAPI) int\u00e8gre des param\u00e8tres tenant compte des d\u00e9gradations optiques, permettant une validation de la couche physique ind\u00e9pendante des fournisseurs et contribuant directement \u00e0 la sp\u00e9cification TAPI v2.4.0. La plateforme int\u00e8gre un cadre de gestion de r\u00e9seau en tant que code (NMaC) qui applique les principes GitOps pour automatiser la gestion du cycle de vie du plan de contr\u00f4le, d\u00e9montr\u00e9e par une migration transparente d’architectures r\u00e9seau partiellement \u00e0 enti\u00e8rement d\u00e9sagr\u00e9g\u00e9es. Deuxi\u00e8mement, deux nouvelles fonctions de contr\u00f4le bas\u00e9es sur l’apprentissage automatique r\u00e9pondent \u00e0 des d\u00e9fis op\u00e9rationnels critiques : un apprentissage par renforcement convolutionnel sur graphes et multi-agents pour le routage et l’allocation de spectre (MAGC-RSA) emploie une prise de d\u00e9cision distribu\u00e9e avec un apprentissage centralis\u00e9 pour optimiser l’allocation des ressources, tandis qu’un cadre de classification de la qualit\u00e9 de transmission par ensemble et adaptatif \u00e0 la d\u00e9rive (DAEQoT) maintient des performances robustes de classification de la qualit\u00e9 de transmission (QoT) dans des conditions de r\u00e9seau changeantes, gr\u00e2ce \u00e0 des m\u00e9canismes dynamiques d’adaptation. Troisi\u00e8mement, un cadre d’apprentissage automatique compositionnel (CMLF) permet la d\u00e9couverte, l’assemblage et l’orchestration automatis\u00e9s de services d’apprentissage automatique distribu\u00e9s, facilitant la cr\u00e9ation de flux de travail d’apprentissage automatique complexes et r\u00e9utilisables. Le cadre est valid\u00e9 par un cas d’usage de RMSA compositionnel combinant la pr\u00e9vision de s\u00e9ries temporelles avec un agent d’apprentissage par renforcement profond sensible \u00e0 la fragmentation. La recherche d\u00e9montre la faisabilit\u00e9 pratique par des simulations approfondies et des impl\u00e9mentations de preuve de concept sur des \u00e9quipements optiques Nokia, montrant des am\u00e9liorations significatives de la probabilit\u00e9 de blocage, de l’efficacit\u00e9 spectrale et de l’automatisation op\u00e9rationnelle. Ces contributions posent les bases de r\u00e9seaux optiques v\u00e9ritablement autonomes, capables d’auto-optimisation, d’auto-r\u00e9paration et de gestion intelligente des ressources, r\u00e9pondant ainsi aux besoins critiques des infrastructures de transport de nouvelle g\u00e9n\u00e9ration. The explosive growth of data traffic and the increasing complexity of optical transport networks have created an urgent need for more flexible, automated, and intelligent network control architectures. Traditional monolithic, vendor-locked optical systems can no longer meet the demands for rapid service deployment, cost-effective scaling, and autonomous network management. This dissertation addresses these challenges by developing a comprehensive framework for infrastructure control and end-to-end service orchestration in future transport networks through the convergence of software-defined networking (SDN), network disaggregation, and machine learning-driven intelligence. The research presents three primary contributions that collectively enable autonomous optical network operation. First, a cloud-native Open and Disaggregated Optical Transport Control Platform is developed, featuring a Transport SDN Controller (T-SDNC) built on OpenDaylight with microservice architecture. A critical enhancement to the Transport API (TAPI) standard integrates optical impairment-aware parameters, enabling vendor-agnostic physical layer validation and contributing directly to the TAPI v2.4.0 specification. The platform incorporates a Network Management as Code (NMaC) framework that applies GitOps principles to automate control plane lifecycle management, demonstrated through seamless migration from partially to fully disaggregated network architectures. Second, two novel ML-driven control functions address critical operational challenges: Multi-Agent Graph Convolutional Reinforcement Learning for Routing and Spectrum Assignment (MAGC-RSA) employs distributed decision-making with centralized learning to optimize resource allocation, while a Drift-Adaptive Ensemble-based Quality of Transmission Classification (DAEQoT) framework maintains robust Quality of Transmission (QoT) classification performance under changing network conditions through dynamic drift-adaptation mechanisms. Third, a Compositional Machine Learning Framework (CMLF) enables automated discovery, stitching, and orchestration of distributed ML services, facilitating the creation of complex, reusable ML workflows for network automation. The framework is validated through a compositional Routing, Modulation and Spectrum Assignment use-case combining time-series forecasting with fragmentation-aware deep reinforcement learning agent. The research demonstrates practical feasibility through extensive simulations and proof-of-concept implementations on Nokia optical equipment, showing significant improvements in blocking probability, spectral efficiency, and operational automation. These contributions establish a foundation for truly autonomous optical networks capable of self-optimization, self-healing, and intelligent resource management, addressing the critical needs of next-generation transport infrastructure.<\/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 Quang Huy TRAN 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":[286,402],"tags":[],"class_list":["post-6984","post","type-post","status-publish","format-standard","hentry","category-fractualites-ennews-fr","category-seminaires-ness-2013-fr","entry"],"_links":{"self":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/6984","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=6984"}],"version-history":[{"count":1,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/6984\/revisions"}],"predecessor-version":[{"id":6985,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/posts\/6984\/revisions\/6985"}],"wp:attachment":[{"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/media?parent=6984"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/categories?post=6984"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/samovar.telecom-sudparis.eu\/index.php\/wp-json\/wp\/v2\/tags?post=6984"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
ID de r\u00e9union: 934 6100 3734<\/a>
Code secret: 647963<\/p>\n\n\n\n
M. Dominique VERCHERE<\/strong>, Ing\u00e9nieur de recherche, Nokia Bell Labs, FRANCE – CoDirecteur de th\u00e8se
M. Pascal LORENZ<\/strong>, Professeur, Universit\u00e9 de Haute-Alsace, FRANCE – Rapporteur
M. Marco RUFFINI<\/strong>, Professeur, University of Dublin, Trinity College, Ireland, IRLANDE – Rapporteur
M. Marc CARTIGNY<\/strong>, EXPERT SCIENTIFIQUE ET TECHNIQUE – EXPERT R&D, AIRBUS DEFENSE AND SPACE, FRANCE – Examinateur
M. Khalil DRIRA<\/strong>, Professeur, LAAS-CNRS, FRANCE – Examinateur<\/p>\n\n\n\n\u00ab Contr\u00f4le d\u2019Infrastructure et Orchestration de Services de bout-en-bout pour les R\u00e9seaux Optiques du Futur \u00bb<\/h2>\n\n\n\n
pr\u00e9sent\u00e9 par Monsieur Quang Huy TRAN<\/h2>\n\n\n\n
Abstract :<\/strong><\/p>\n\n\n\n