DC Field | Value | Language |
dc.contributor.author | BENSALEM, AKram | - |
dc.date.accessioned | 2023-10-18T13:18:33Z | - |
dc.date.available | 2023-10-18T13:18:33Z | - |
dc.date.issued | 2023 | - |
dc.identifier.uri | https://repository.esi-sba.dz/jspui/handle/123456789/556 | - |
dc.description | Encadreur : M BELFEDHAL Alaa Eddine / Co-Encadreur : M LALLET Julien | en_US |
dc.description.abstract | Abstract :
In today’s world, the demand for efficient and reliable cloud computing systems is
increasing exponentially. The emergence of edge cloud systems has brought about a new
level of convenience and accessibility for various industries and applications. However,
managing data workloads efficiently in edge cloud systems, especially for connected cars,
can be a challenging task. To address this issue, we introduce a new method for cache
management prefetching that uses visibility graphs to handle time series data workloads
effectively.
Our approach involves forecasting future data from the storage sources and preloading
this data into the cache in the edge cloud. Doing so reduces data retrieval time and
boosts overall system performance. The visibility graph based method helps to manage
the workload in real time, which is critical in many use cases. By adopting this approach,
we can guarantee that the system operates efficiently and effectively, minimizing the risk
of delays caused by the system.
We tested our approach through simulations and experiments and found that it significantly
improves the performance of edge cloud systems. The study proves that our
approach efficiently handles time series data workloads in edge cloud systems, speciőcally
for connected cars. As technology progresses and demands for cloud computing systems
continue to grow, őnding innovative and efficient solutions to manage data workloads is
crucial. Our approach offers a promising solution that can help address this challenge and
pave the way for a more efficient and reliable cloud computing system.
Although our initial motivation was intelligent transportation, this solution can also apply
to other systems based on the edge-cloud concept. Our approach can beneőt real-time
data processing applications in various industries. Moreover, our solution is simple to
integrate into existing edge cloud systems, making it a practical and valuable tool for
organizations looking to enhance their data processing capabilities.
In conclusion, our solution offers a powerful combination of simplicity, speed, and scalability.
With our approach to edge-cloud computing, organizations can unlock the full
potential of real time data processing and drive meaningful results across various industries.***
Résumé :
Dans le monde actuel, la demande pour des systèmes de cloud computing (cloud systems)
efficaces et őables croît de façon exponentielle. L’émergence du cloud computing de
bord (edge computing) offre un nouveau niveau de commodité et d’accessibilité à diverses
industries et applications. Toutefois, gérer efficacement les charges de travail des données
(data workloads), en particulier pour les voitures connectées (connected cars), demeure un
déő. Pour répondre à cette problématique, nous proposons une méthode innovante pour
la gestion du cache prefetching en utilisant des graphes de visibilité (visibility graphs)
pour traiter adéquatement les charges de travail des données séries chronologiques (time
series data).
Notre stratégie prévoit les données futures (future data) depuis leurs sources de stockage
pour les précharger (preload) dans le cache du cloud de bord (edge cloud). Ceci
minimise le temps de récupération des données et optimise les performances globales du
système. La méthode basée sur les graphes de visibilité (visibility graph-based method)
permet de gérer la charge de travail en temps réel, aspect essentiel pour de nombreux cas
d’utilisation. Grâce à cette méthode, nous garantissons une performance ŕuide et efficace
du système, réduisant ainsi les risques de latences.
Aőn d’évaluer notre méthode, nous avons réalisé des simulations et des tests. Les résultats
démontrent une nette amélioration des performances du cloud computing de bord
(edge cloud computing). Ils indiquent que notre stratégie est parfaite pour gérer les
charges de travail des données séries chronologiques (time series data workload), en particulier
pour les voitures connectées (connected cars). Alors que la technologie évolue | en_US |
dc.language.iso | en | en_US |
dc.subject | Big Data | en_US |
dc.subject | Caching | en_US |
dc.subject | Visibility Graph | en_US |
dc.subject | Edge Cloud | en_US |
dc.subject | Time Series | en_US |
dc.title | Efficient Time Series Data Management in the Edge Cloud based on Visibility Graph | en_US |
dc.type | Thesis | en_US |
Appears in Collections: | Ingénieur
|