Dissertations / Theses on the topic 'Découpage du réseau dans la 5G'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 15 dissertations / theses for your research on the topic 'Découpage du réseau dans la 5G.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
Chang, Chia-Yu. "Cloudification and Slicing in 5G Radio Access Network." Electronic Thesis or Diss., Sorbonne université, 2018. https://accesdistant.sorbonne-universite.fr/login?url=https://theses-intra.sorbonne-universite.fr/2018SORUS293.pdf.
Full textOver the past few decades, the continuing growth of network statistics requires a constantly evolving technology. Therefore, a natural question arises in our minds: what will 5G be? To answer this question, the 5G architecture must be designed with a certain level of flexibility through the integration of softwarization and virtualization principles. Therefore, we can see that 5G will provide a paradigm shift beyond radio access technology in order to establish an agile and sophisticated communication system. The network can be used efficiently and independently by creating multiple logically separated spaces, called network slices. In addition, each logical network can deploy its network functions in a flexible cloud environment. To this end, the goal of this thesis is to study these two techniques: (a) Cloud-RAN and (b) RAN splitting. In the first part, our focus is on the C-RAN concept, in which monolithic base stations are replaced by (1) distributed radio elements and (2) centralized pools for baseband processing units. The C-RAN notion is still confronted with stringent capacity and latency requirements of the fronthaul interface that connects the distributed remote radio unit to the centralized baseband processing unit. In the second part, we focus on RAN cutting not only to allow different levels of isolation and sharing at each slice of network, but also to customize the control plane, user plane and control logic. Therefore, we provide a flexible runtime environment for the "RAN Runtime" slicing system to host service instances on each of the underlying RAN modules
Bakri, Sihem. "Towards enforcing network slicing in 5G networks." Electronic Thesis or Diss., Sorbonne université, 2021. http://www.theses.fr/2021SORUS067.
Full textThe current architecture “one size fits all” of 4G network cannot support the next-generation 5G heterogeneous services criteria. Therefore, research around 5G aims to provide more adequate architectures and mechanisms to deal with this purpose. The 5G architecture is envisioned to accommodate the diverse and conflicting demands of services in terms of latency, bandwidth, and reliability, which cannot be sustained by the same network infrastructure. In this context, network slicing provided by network virtualization allows the infrastructure to be divided into different slices. Each slice is tailored to meet specific service requirements allowing different services (such as automotive, Internet of Things, etc.) to be provided by different network slice instances. Each of these instances consists of a set of virtual network functions that run on the same infrastructure with specially adapted orchestration. Three main service classes of network slicing have been defined by the researchers as follows: Enhanced Mobile Broadband (eMBB), massive Machine Type Communication (mMTC), and ultra-Reliable and Low-Latency Communication (uRLLC). One of the main challenges when it comes to deploying Network Slices is slicing the Radio Access Network (RAN). Indeed, managing RAN resources and sharing them among Network Slices is an increasingly difficult task, which needs to be properly designed. This thesis proposes solutions that aim to improve network performance, and introduce flexibility and greater utilization of network resources by accurately and dynamically provisioning the activated network slices with the appropriate amounts of resources to meet their diverse requirements
Eido, Souheir. "Contrôle de la mobilité dans un réseau d'opérateur convergé fixe-mobile." Thesis, Ecole nationale supérieure Mines-Télécom Atlantique Bretagne Pays de la Loire, 2017. http://www.theses.fr/2017IMTA0025/document.
Full textFixed and mobile networks are currently experiencing a dramatic growth in terms of data traffic, mainly driven by video content distribution. Telecoms operators are thus considering de-centralizing content distribution architecture for future Fixed and Mobile Converged (FMC) network architectures. This decentralization, together with a distributed mobile EPC, would be used for future 5G networks. Mobile data offloading, in particular SIPTO approaches, already represent a good implementation model for 5G network as it allows the use of distributed IP edges to offload Selected IP traffic off the currently centralized mobile core network. However, in some cases, SIPTO does not support session continuity during users' mobility. This is due to the fact that user's mobility may imply packet gateway (PGW) relocation and thus a modification of the UE's IP address.This PhD thesis first quantifies the gain, in terms of bandwidth demands on various network portions, brought by the generalized use of mobile traffic offloading. A state of art of existing mobile data offloading solutions is presented, showing that none of the existing solutions solve the problem of session continuity for long-lived sessions. This is why, in the context of future FMC mobile network architectures, the PhD thesis proposes solutions to provide seamless mobility for users relying on SIPTO with the help of Multipath TCP (MPTCP). 3GPP standards are not modified, as session continuity is ensured by end-points. Lastly, the proposed solutions are mapped on different architecture options considered for future FMC networks
Matoussi, Salma. "User-Centric Slicing with Functional Splits in 5G Cloud-RAN." Electronic Thesis or Diss., Sorbonne université, 2021. https://accesdistant.sorbonne-universite.fr/login?url=https://theses-intra.sorbonne-universite.fr/2021SORUS004.pdf.
Full text5G Radio Access Network (RAN) aims to evolve new technologies spanning the Cloud infrastructure, virtualization techniques and Software Defined Network capabilities. Advanced solutions are introduced to split the RAN functions between centralized and distributed locations to improve the RAN flexibility. However, one of the major concerns is to efficiently allocate RAN resources, while supporting heterogeneous 5G service requirements. In this thesis, we address the problematic of the user-centric RAN slice provisioning, within a Cloud RAN infrastructure enabling flexible functional splits. Our research aims to jointly meet the end users’ requirements, while minimizing the deployment cost. The problem is NP-hard. To overcome the great complexity involved, we propose a number of heuristic provisioning strategies and we tackle the problem on four stages. First, we propose a new implementation of a cost efficient C-RAN architecture, enabling on-demand deployment of RAN resources, denoted by AgilRAN. Second, we consider the network function placement sub-problem and propound a new scalable user-centric functional split selection strategy named SPLIT-HPSO. Third, we integrate the radio resource allocation scheme in the functional split selection optimization approach. To do so, we propose a new heuristic based on Swarm Particle Optimization and Dijkstra approaches, so called E2E-USA. In the fourth stage, we consider a deep learning based approach for user-centric RAN Slice Allocation scheme, so called DL-USA, to operate in real-time. The results obtained prove the efficiency of our proposed strategies
Zeng, Xuan. "Vers une mobilité transparente dans le réseau ICN : connectivité, sécurité, et fiabilité." Thesis, Sorbonne université, 2018. http://www.theses.fr/2018SORUS046/document.
Full textWith the proliferation of mobile devices, mobility becomes a requirement and a compelling feature for 5G. However, despite tremendous efforts in the last 2 decades to enable mobility in IP network, the solutions are mostly anchor-based and inefficient. In this context, Information-Centric networking (ICN) is proposed. While ICN has some native support of mobility, other architectural challenges remain unsolved to achieve seamless mobility. The thesis explores 3 main challenges of such and contributes novel solutions. First, to solve producer mobility, MapMe, a micro mobility management protocol supporting latency sensitive traffic is proposed. MAP-Me is anchorless and preserves key ICN benefits. Simulation results show that MAP-Me outperforms existing work in user performance while retaining low network overheads in various network conditions. Second, we investigate security in producer mobility. We focus on prefix hijacking attack, which is a basis of several attacks. To prevent prefix hijacking, we propose a light-weight and distributed prefix attestation protocol based on hash-chaining. First results show significant improvement in verification overhead. It is resistant to replay-based prefix hijacking. Finally, additional transport-layer mechanisms are needed in mobile ICN. To this aim, we investigate alleviating the adverse effect of wireless/mobility loss on congestion control. We propose WLDR and MLDR for in-network loss detection and recovery to facilitate congestion control. Simulation results show a significant reduction in flow completion time (up to 20%)
Manini, Malo. "Allocation de ressources et ordonnancement dans les réseaux de 5ème génération." Thesis, Rennes 1, 2021. http://www.theses.fr/2021REN1S003.
Full textThe increasing number of wireless network users and the diversification of their usage call for an evolution of the resource management methods. This thesis is based on 5G resource allocation techniques. In regular wireless networks, cells are managed independently. In this context, we propose a resource allocation algorithm with the aim of fairly guaranteeing the best service to users. When the cell charge is low enough to guarantee a sufficient quality of service, the algorithm redirects dynamically its priorities towards energy saving. This behavior allows to obtain an efficient compromise between capacity and energy consumption at different charge levels. In order to extend network capacity, the adding of new cells allows to broaden the available bandwidth and to reduce the distance induced signal attenuation. We present an algorithm of user repartition in a multi-cell context, which intervenes before the resource allocation stage. A user can be covered by several cells using different frequencies, thus its repartition will have strong repercussions on the cell charge balance and on the general quality of service in the system. The maximal number of cells in a sector is limited by its geographical environment. The Massive-MIMO allows to increase the cell functionalities while allowing a better energy directivity, and thus adding the spatial component to the resource allocation. We propose a new indicator evaluating the spatial compatibility of users based on past allocations. Once integrated in an allocation algorithm, it takes advantage of the superior capacities of Massive-MIMO
Sapountzis, Nikolaos. "Optimisation au niveau réseau dans le cadre des réseaux hétérogènes nouvelle génération." Electronic Thesis or Diss., Paris, ENST, 2016. http://www.theses.fr/2016ENST0082.
Full textBy 2016, it is well-known that mobile networking has dominated our lives. We use our mobile cell phones for almost everything: from social networking to streaming, finding accommodation or banking. Nevertheless, it seems that operators have not understood yet this domination, since their networks consist of nodes that: (i) suffer from enormous load fluctuations, (ii) waste their resources, and (iii) are blamed to be a major energy-killer worldwide. Such shortcomings hurt: load-balancing, spectral and energy efficiency, respectively. The goal of this dissertation is to carefully study these efficiencies and achieve a good trade-off between them for future mobile 5G heterogeneous networks (HetNets). Towards this direction, we firstly focus on (i) the user and traffic differentiation, emerging from the MTC and IoT applications, and (ii) the RAN. Specifically, we perform appropriate modeling, performance analysis and optimization for a family of objectives, using tools mostly coming from (non) convex optimization, probability and queueing theory. Our initial consideration is on network-layer optimizations (e.g. studying the user association problem). Then, we analytically show that cross-layer optimization is key for the success of future HetNets, as one needs to jointly study other problems coming from the layers below (e.g. the TDD allocation problem from the MAC, or the cross-interference management from the PHY) to avoid performance degradation. Finally, we add the backhaul network into our framework, and consider additional constraints related to the backhaul capacity, backhaul topology, as well as the problem of backhaul TDD allocation
Mouawad, Nadia. "SDN based Mobility Management and Quality of Service Provisioning for 5G Vehicular Networks." Thesis, université Paris-Saclay, 2020. http://www.theses.fr/2020UPASV003.
Full textVehicle to everything (V2X), including vehicle-to-vehicle (V2V) and vehicle-to-infrastructure(V2I), is the umbrella for the vehicular communication system, where active road safety, infotainment and traffic management messages are transmitted over high-bandwidth, low-latency, high-reliability links, paving the way to fully autonomous driving. The ultimate objective of next generation V2X communication systems is enabling accident-free cooperative driving that uses the available roadway efficiently. To achieve this goal, the communication system will need to enable a diverse set of use cases, each with a specific set of requirements.The main use case categories requirements analysis, specifically the critical realtime applications, points out the need for an efficient V2X system design that could fulfill the network performance. The Fifth Generation (5G) technology, with its provisioned QoS features in terms of high capacity and low latency, is advocated as a prominent solution to cope with the firm requirements imposed by V2X applications.In this multifaceted vehicular 5G ecosystem, diverse communication technologies are envisioned, spanning from IEEE 802.11p, LTE, LTE-V to vehicular visible light communications. Therefore, the heterogeneity of radio access technologies will raise a concern regarding the seamless mobility management and the quality of service guarantee.This thesis provides a novel mobility management scheme devised for 5G vehicular networks based on the emerging Software Defined Networking (SDN) technology.SDN provides network programmability that strives to achieve an efficient network resource allocation and mobility management.Our research work tackles three objectives. At a first stage, we design a software defined vehicular network topology. On the top of this topology, we implement twoSDN applications, namely Network Selection Application and Mobility Management Application. The proposed architecture is enhanced by a controller placement solution that aims at reducing communication latency. Moreover, a special concern is devoted to design a SDN road active safety application that controls speed traps placement. The proposed application aims at reducing accidents rate which is a main purpose of future Intelligent Transportation System.The second objective of this thesis tackles the mobility management problem. This is achieved by implementing SDN mobility related applications on the top of the adopted network topology. The first application is dedicated to solve the network selection problem; it aims at mapping running V2X sessions to the corresponding technology. The second application is conceived to solve the handover procedure; this is achieved using packets duplication and introducing an efficient routing algorithm.The third thesis objective is focused on QoS provisioning for V2X communications
Zeng, Xuan. "Vers une mobilité transparente dans le réseau ICN : connectivité, sécurité, et fiabilité." Electronic Thesis or Diss., Sorbonne université, 2018. http://www.theses.fr/2018SORUS046.
Full textWith the proliferation of mobile devices, mobility becomes a requirement and a compelling feature for 5G. However, despite tremendous efforts in the last 2 decades to enable mobility in IP network, the solutions are mostly anchor-based and inefficient. In this context, Information-Centric networking (ICN) is proposed. While ICN has some native support of mobility, other architectural challenges remain unsolved to achieve seamless mobility. The thesis explores 3 main challenges of such and contributes novel solutions. First, to solve producer mobility, MapMe, a micro mobility management protocol supporting latency sensitive traffic is proposed. MAP-Me is anchorless and preserves key ICN benefits. Simulation results show that MAP-Me outperforms existing work in user performance while retaining low network overheads in various network conditions. Second, we investigate security in producer mobility. We focus on prefix hijacking attack, which is a basis of several attacks. To prevent prefix hijacking, we propose a light-weight and distributed prefix attestation protocol based on hash-chaining. First results show significant improvement in verification overhead. It is resistant to replay-based prefix hijacking. Finally, additional transport-layer mechanisms are needed in mobile ICN. To this aim, we investigate alleviating the adverse effect of wireless/mobility loss on congestion control. We propose WLDR and MLDR for in-network loss detection and recovery to facilitate congestion control. Simulation results show a significant reduction in flow completion time (up to 20%)
Arora, Sagar. "Cloud Native Network Slice Orchestration in 5G and Beyond." Electronic Thesis or Diss., Sorbonne université, 2023. http://www.theses.fr/2023SORUS278.
Full textNetwork Function Virtualization (NFV) is the founding pillar of 5G Service Based Architecture. It has the potential to revolutionize the future mobile communication generations. NFV started long back in 2012 with Virtual-Machine (VM) based Virtual Network Functions (VNFs). The use of VMs raised multiple questions because of the compatibility issues between VM hypervisors and their high resource consumption. This made containers to be an alternative network function packaging technology. The lightweight design of containers improves their instantiation time and resource footprints. Apart from network functions, containerization can be a promising enabler for Multi-access Edge Computing (MEC) applications that provides a home to low-latency demanding services. Edge computing is one of the key technology of the last decade, enabling several emerging services beyond 5G (e.g., autonomous driving, robotic networks, Augmented Reality (AR)) requiring high availability and low latency communications. The resource scarcity at the edge of the network requires technologies that efficiently utilize computational, storage, and networking resources. Containers' low-resource footprints make them suitable for designing MEC applications. Containerization is meant to be used in the framework of cloud-native application design fundamentals, loosely coupled microservices-based architecture, on-demand scalability, and high resilience. The flexibility and agility of containers can certainly benefit 5G Network Slicing that highly relies on NFV and MEC. The concept of Network slicing allows the creation of isolated logical networks on top of the same physical network. A network slice can have dedicated network functions or its network functions can be shared among multiple slices. Indeed, network slice orchestration requires interaction with multiple technological domain orchestrators, access, transport, core network, and edge computing. The paradigm shift of using cloud-native application design principles has created challenges for legacy orchestration systems and the ETSI NFV and MEC standards. They were designed for handling virtual machine-based network functions, restricting them in their approach to managing a cloud-native network function. The thesis examines the existing standards of ETSI NFV, ETSI MEC, and network service/slice orchestrators. Aiming to overcome the challenges around multi-domain cloud-native network slice orchestration. To reach the goal, the thesis first proposes MEC Radio Network Information Service (RNIS) that can provide radio information at the subscriber level in an NFV environment. Second, it provides a Dynamic Resource Allocation and Placement (DRAP) algorithm to place cloud-native network services considering their cost and availability matrix. Third, by combining NFV, MEC, and Network Slicing, the thesis proposes a novel Lightweight edge Slice Orchestration framework to overcome the challenges around edge slice orchestration. Fourth, the proposed framework offers an edge slice deployment template that allows multiple possibilities for designing MEC applications. These possibilities were further studied to understand the impact of the microservice design architecture on application availability and latency. Finally, all this work is combined to propose a novel Cloud-native Lightweight Slice Orchestration (CLiSO) framework extending the previously proposed Lightweight edge Slice Orchestration (LeSO) framework. In addition, the framework offers a technology-agnostic and deployment-oriented network slice template. The framework has been thoroughly evaluated via orchestrating OpenAirInterface container network functions on public and private cloud platforms. The experimental results show that the framework has lower resource footprints than existing orchestrators and takes less time to orchestrate network slices
Saliba, Danielle. "WIFI Integration with LTE in the Roadmap of 5G Networks." Thesis, Ecole nationale supérieure Mines-Télécom Atlantique Bretagne Pays de la Loire, 2019. http://www.theses.fr/2019IMTA0168.
Full textFollowing the continuous increase of the mobile data traffic, the 5G technology has been introduced to offer additional capacity and higher data rate. WiFi APs integration with mobile networks are considered as potential candidate towards the heterogenous networks. We first propose in this thesis an algorithm for the estimation of the WiFi physical channels load through the observation of the non-overlapped channels and estimating as a result the load of the entire physical channels. Then, in our second proposed algorithm, we propose to dimension the WiFi network to offload LTE network and transfer the users that consume the high level of transmission power. The algorithm calculates the minimum needed number of APs that will support the extra offloaded LTE traffic. Finally, we evaluate the benefit of the cooperation by estimating the profit share of LTE and WiFi. We calculate for each player the profit using a coalition game concept based on Shapley value
Bekkar, Mohammed. "Formation de voies hybride analogique-numérique pour la réduction d'interférences dans les réseaux cellulaires de nouvelle génération." Thesis, Université Grenoble Alpes, 2020. http://www.theses.fr/2020GRALT007.
Full textBeamforming is a signal processing method used in antenna arrays, allowing to enhance directions of emission or reception of signals by controlling the different elements.In mobile networks especially, it allows interference reduction in base stations.Its full digital impementation is limited by energy consumption and cost when increasing the number of antennas.As a response, hybrid analog-digital implementation could be used to reduce the number of radiofrequency (RF) chains as well as the number of analog-to-digital converters.In this implementation, the analog stage could be realised using different types of devices (phase shifters, amlifiers/attenuators, variable impedances) and with a variable connectivity to the antenna array.Nevertheless, if we want to keep a simple RF circuitry by using phase shifters only to tune the analog beamformer, the problem of optimising these weights becomes non-convex.The current works on small cell networks show that the interference between base station is one of the limiting factors of the coverage and the datarate.Furthermore, in a full digital implementation, the presence of strong blockers leads to analog-to-digital converters saturation or desensitization.The purpose of this work is the study of hybrid beamforming with phase-only implementation, as well as to propose an algorithm to compute the beamforming matrices, to reduce the received interference in a small cell.After a description and a state-of-the-art, we preliminarily proposed an interference characterization using an algebraic angle between the signals of interest vectors and the interference vectors, which allowed us to obtain a lower bound on the SINR performance of the optimal beamformer.We have then proposed a sub-optimal solution of hybrid phase-only beamforming, which when using an infinite resolution digitization, has a low loss as compared to a solution using modulus and phase.Secondly, we introduced an analog-to-digital converter model, which allowed us to bring out the limitations of the first appproach as well as of the full digital implementation, in the presence of strong blockers.Afterwards, we proposed an optimisation algorithm of the analog stage, based on a semidefinite relaxation.The peroformance of this algorithm, in terms of SINR and sumrate are close to the benchmark with full degree of freedom, modulus and phase.In comparison, the performance state-of-the-art tested solutions using non-convex cost function are lower and depend on initialization point
Morcos, Mira. "Auction-based dynamic resource orchestration in cloud-based radio access networks." Thesis, Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLL003.
Full textNetwork densification using small cells massively deployed over the macro-cell areas, represents a promising solution for future 5G mobile networks to cope with mobile traffic increase. In order to simplify the management of the heterogeneous Radio Access Network (RAN) that results from the massive deployment of small cells, recent research and industrial studies have promoted the design of novel centralized RAN architectures termed as Cloud-RAN (C-RAN), or Virtual RAN (V-RAN), by incorporating the benefits of cloud computing and Network Functions Virtualization (NFV). The DynaRoC project aims at (1) developing a theoretical framework of resource orchestration for C-RAN and deriving the fundamental performance limits as well as the tradeoffs among various system parameters, and (2) designing dynamic resource orchestration mechanisms based on the theoretical findings to achieve a desired performance balance, by taking into account various design challenges. The PhD student will investigate innovative resource optimization mechanisms to foster the deployment of C-RANs, improving their performance exploiting the enabling Network Functions Virtualization technology
Boutiba, Karim. "On enforcing Network Slicing in the new generation of Radio Access Networks." Electronic Thesis or Diss., Sorbonne université, 2024. https://accesdistant.sorbonne-universite.fr/login?url=https://theses-intra.sorbonne-universite.fr/2024SORUS003.pdf.
Full textThe emerging 5G networks and beyond promise to support novel use cases such as immersive holographic communication, Internet of Skills, and 4D Interactive mapping [usecases]. These use cases require stringent requirements in terms of Quality of Service (QoS), such as low latency, high Downlink (DL)/Uplink (UL) throughput and low energy consumption. The 3rd Generation Partnership Project (3GPP) specifications introduced many features in 5G New Radio (NR) to improve the physical efficiency of 5G to meet the stringent and heterogeneous requirements of beyond 5G services. Among the key 5G NR features, we can mention the numerology, BandWidth Part (BWP), dynamic Time Duplex Division (TDD) and Connected-mode Discontinuous Reception (C-DRX). However, the specifications do not provide how to configure the next Generation Node B (gNB)/User Equipment (UE) in order to optimize the usage of the 5G NR features. We enforce the 5G NR features by applying Machine Learning (ML), particularly Deep Reinforcement Learning (DRL), to fill this gap. Indeed, Artificial Intelligence (AI)/ML is playing a vital role in communications and networking [1] thanks to its ability to provide a self-configuring and self-optimizing network.In this thesis, different solutions are proposed to enable intelligent configuration of the Radio Access Network (RAN). We divided the solutions into three different parts. The first part concerns RAN slicing leveraging numerology and BWPs. In contrast, the second part tackles dynamic TDD, and the last part goes through different RAN optimizations to support Ultra-Reliable and Low-Latency Communication (URLLC) services.In the first part, we propose two contributions. First, we introduce NRflex, a RAN slicing framework aligned with Open RAN (O-RAN) architecture. NRflex dynamically assigns BWPs to the running slices and their associated User Equipment (UE) to fulfill the slices' required QoS. Then, we model the RAN slicing problem as a Mixed-Integer Linear Programming (MILP) problem. To our best knowledge, this is the first MILP modeling of the radio resource management featuring network slicing, taking into account (i) Mixed-numerology, (ii) both latency and throughput requirements (iii) multiple slices attach per UE (iv) Inter-Numerology Interference (INI). After showing that solving the problem takes an exponential time, we consider a new approach in a polynomial time, which is highly required when scheduling radio resources. The new approach consists of formalizing this problem using a DRL-based solver.In the second part of this thesis, we propose a DRL-based solution to enable dynamic TDD in a single 5G NR cell. The solution is implemented in OAI and tested using real UEs. Then, we extend the solution by leveraging Multi-Agent Deep Reinforcement Learning (MADRL) to support multiple cells, considering cross-link interference between cells.In the last part, we propose three solutions to optimize the RAN to support URLLC services. First, we propose a two-step ML-based solution to predict Radio Link Failure (RLF). We combine Long Short-Term Memory (LSTM) and Support Vector Machine (SVM) to find the correlation between radio measurements and RLF. The RLF prediction model was trained with real data obtained from a 5G testbed. In the second contribution, we propose a DRL-based solution to reduce UL latency. Our solution dynamically allocates the future UL grant by learning from the dynamic traffic pattern. In the last contribution, we introduce a DRL-based solution to balance latency and energy consumption by jointly deriving the C-DRX parameters and the BWP configuration
Esteves, José Jurandir Alves. "Optimization of network slice placement in distributed large-scale infrastructures : from heuristics to controlled deep reinforcement learning." Electronic Thesis or Diss., Sorbonne université, 2021. http://www.theses.fr/2021SORUS325.
Full textThis PhD thesis investigates how to optimize Network Slice Placement in distributed large-scale infrastructures focusing on online heuristic and Deep Reinforcement Learning (DRL) based approaches. First, we rely on Integer Linear Programming (ILP) to propose a data model for enabling on-Edge and on-Network Slice Placement. In contrary to most studies related to placement in the NFV context, the proposed ILP model considers complex Network Slice topologies and pays special attention to the geographic location of Network Slice Users and its impact on the End-to-End (E2E) latency. Extensive numerical experiments show the relevance of taking into account the user location constraints. Then, we rely on an approach called the “Power of Two Choices"(P2C) to propose an online heuristic algorithm for the problem which is adapted to support placement on large-scale distributed infrastructures while integrating Edge-specific constraints. The evaluation results show the good performance of the heuristic that solves the problem in few seconds under a large-scale scenario. The heuristic also improves the acceptance ratio of Network Slice Placement Requests when compared against a deterministic online ILP-based solution. Finally, we investigate the use of ML methods, more specifically DRL, for increasing scalability and automation of Network Slice Placement considering a multi-objective optimization approach to the problem. We first propose a DRL algorithm for Network Slice Placement which relies on the Advantage Actor Critic algorithm for fast learning, and Graph Convolutional Networks for feature extraction automation. Then, we propose an approach we call Heuristically Assisted Deep Reinforcement Learning (HA-DRL), which uses heuristics to control the learning and execution of the DRL agent. We evaluate this solution trough simulations under stationary, cycle-stationary and non-stationary network load conditions. The evaluation results show that heuristic control is an efficient way of speeding up the learning process of DRL, achieving a substantial gain in resource utilization, reducing performance degradation, and is more reliable under unpredictable changes in network load than non-controlled DRL algorithms