DC Field | Value | Language |
dc.contributor.author | LATRECHE, YAssine | - |
dc.date.accessioned | 2024-09-26T09:45:17Z | - |
dc.date.available | 2024-09-26T09:45:17Z | - |
dc.date.issued | 2024 | - |
dc.identifier.uri | https://repository.esi-sba.dz/jspui/handle/123456789/680 | - |
dc.description | Encadreur : M KHALDI Miloud | en_US |
dc.description.abstract | The Internet of Things (IoT) has brought unprecedented internet connectivity
opportunities across various domains with the needs for efficient and
scalable communication technologies. Long Range Wide Area Network (Lo-
RaWAN) stands out for its capability of long-range, low-power consumption
at a cost-effective manner, and it is preferred in several IoT applications.
However, optimizing IoT network performance is still a critical challenge,
especially through appropriate Spreading Factor (SF) allocation. Through
a comprehensive review of the state-of-the-art research of SF optimization
in LoRaWAN networks, this thesis has explored a variety of SF optimization
methodologies and their implications on Packet Delivery Ratio (PDR),
throughput, energy consumption, and network scalability.
In addition to reviewing existing SF optimization methodologies, this thesis
has covered optimization algorithms, such as Genetic Algorithm (GA),
Simulated Annealing (SA), and Differential Evolution (DE), which have
shown promise in outperforming in network performance through SF optimization.
It has also explored surrogate models for rapid optimization
by estimating the performance of complex simulations. These optimization
algorithms and surrogate models offer large improvements in terms of convergence
speed and solution quality. More importantly, they provide us with the
necessary tools to address the challenges of dense IoT environments. These
findings serve as an essential guidance for network planners and operators,
and pave the way for enhanced LoRaWAN deployments. ***
L’Internet des Objets (IoT) a ouvert des opportunit´es sans pr´ec´edent pour
la connectivit´e dans divers domaines, n´ecessitant des technologies de communication
efficaces et ´evolutives. Le r´eseau longue port´ee (LoRaWAN) se
distingue par ses capacit´es `a longue port´ee, sa faible consommation d’´energie
et son coˆut abordable, en faisant un choix privil´egi´e pour de nombreuses
applications IoT. Cependant, l’optimisation des performances du r´eseau, en
particulier par une allocation efficace du facteur d’´etalement (SF), demeure
un d´efi crucial. Cette th`ese m`ene une revue exhaustive des recherches de
pointe sur l’optimisation du SF dans les r´eseaux LoRaWAN, en examinant
diverses m´ethodologies et leurs implications sur le taux de livraison des paquets
(PDR), le d´ebit, la consommation d’´energie et l’´evolutivit´e globale du
r´eseau. En synth´etisant les insights de neuf articles de r´ef´erence, l’´etude
met en ´evidence les complexit´es et les compromis impliqu´es dans l’allocation
du SF, soulignant la n´ecessit´e de disposer d’environnements de simulation
r´ealistes et de strat´egies d’optimisation globales.
En plus de revoir les approches existantes, cette th`ese explore des algorithmes
d’optimisation avanc´es tels que l’algorithme g´en´etique (GA), le recuit
simul´e (SA) et l’´evolution diff´erentielle (DE), qui ont montr´e leur potentiel
pour am´eliorer les performances du r´eseau grˆace `a une allocation optimale
du SF. L’´etude examine ´egalement le potentiel des mod`eles substituts pour
acc´el´erer le processus d’optimisation en approximant les performances de
simulations complexes. Ces algorithmes et mod`eles sophistiqu´es offrent des
am´eliorations significatives en termes de vitesse de convergence et de qualit´e
des solutions, en faisant des outils pr´ecieux pour relever les d´efis des environnements
IoT denses et h´et´erog`enes. Les r´esultats fournissent des conseils
pr´ecieux pour les planificateurs et les op´erateurs de r´eseaux, ouvrant la voie
`a des d´eploiements LoRaWAN am´elior´es capables de r´epondre aux exigences
des environnements IoT de plus en plus denses et h´et´erog`enes. | en_US |
dc.language.iso | en | en_US |
dc.subject | Internet of Things (IoT) | en_US |
dc.subject | Long Range Wide Area Network (LoRaWAN) | en_US |
dc.subject | Spreading Factor (SF) Allocation | en_US |
dc.subject | Packet Delivery Ratio (PDR) | en_US |
dc.subject | Network Optimization | en_US |
dc.subject | Simulation Environments | en_US |
dc.subject | Genetic Algorithm (GA) | en_US |
dc.subject | Simulated Annealing (SA) | en_US |
dc.subject | Differential Evolution (DE) | en_US |
dc.subject | Surrogate Models | en_US |
dc.subject | Machine Learning | en_US |
dc.title | Optimization methods for LPWAN networks | en_US |
dc.type | Thesis | en_US |
Appears in Collections: | Master
|