Information
Thesis supervisor :
Start year :
End 2017
University :
Université Constantine 2
Viva defense
Viva defense date :
1 February 2022
Jury members :
Salim Chikhi
Djamel Eddine Saidouni
Abdelhafid Zitouni
Ghalem Belalem
Salim Bitam
Details
Titre :
Elaboration d'une approche diagnostic des systèmes opérant dans les environnements des objets connectés

Résumé :

Internet of Things paradigm changed the vision of the distributed computer, the traditional internet, and economy to create a future internet. On the basis of this paradigm, a multidimension vision was established with a new form of communication that includes anywhere, everywhere at any time, which poses two major challenges: communication of things and the management of them. Maintaining the healthy state of the system and its elements is crucial to deliver high services quality that conforms to the user’s requirements. However, the IoT service itself should be first available, aware of its context, perceptible of its quality, and able to be servitisation to be delivered correctly. Ensuring those features under any conditions, detect and recover from anomalies, its own a failure diagnosis problem. Yet, the higher distribution, the IoTcontexts dynamicity, and the hyper-connectivity of things complicate the problem. This requires the integration of other paradigms and computer properties, such that, the cloud computer and the autonomic computer. knowing how to combine them to benefit and not complicate the failure diagnostic problem remains to be challenging in itself, where a few literature solvers use this combination to face this problem to ensure the IoT service correctness.

This thesis aims to define a multi-dimensional diagnostic solution that fits the system characteristics regardless of the heterogeneity of the attached things and targets the problem from different angles. Relying on the literature review in this thesis, three major problems are selected to deal along with, which are service availability, contextualisation, and servitisation problems in the IoT paradigm. Considering these three problems and having an aspect of their solving in a transparent way, we map the state-of-the-art solver based on their integrated approaches, their mechanisms, and techniques in order to define the limitation based on each axis. Taking into count those findings, five contributions are proposed using the distributed algorithms and the autonomic computer to answer the research questions, the closure (safety) and the convergence (liveness) properties will be defined and proved satisfied by two proposed protocols. In addition, to highlight the importance of these protocols, we present an application in the context of IoT solutions.

Title:
An Elaboration of a Diagnostic Approach that Operates in a Connected Objects Environment

Abstract:

Internet of Things paradigm changed the vision of the distributed computer, the traditional internet, and economy to create a future internet. On the basis of this paradigm, a multidimension vision was established with a new form of communication that includes anywhere, everywhere at any time, which poses two major challenges: communication of things and the management of them. Maintaining the healthy state of the system and its elements is crucial to deliver high services quality that conforms to the user’s requirements. However, the IoT service itself should be first available, aware of its context, perceptible of its quality, and able to be servitisation to be delivered correctly. Ensuring those features under any conditions, detect and recover from anomalies, its own a failure diagnosis problem. Yet, the higher distribution, the IoTcontexts dynamicity, and the hyper-connectivity of things complicate the problem. This requires the integration of other paradigms and computer properties, such that, the cloud computer and the autonomic computer. knowing how to combine them to benefit and not complicate the failure diagnostic problem remains to be challenging in itself, where a few literature solvers use this combination to face this problem to ensure the IoT service correctness.

This thesis aims to define a multi-dimensional diagnostic solution that fits the system characteristics regardless of the heterogeneity of the attached things and targets the problem from different angles. Relying on the literature review in this thesis, three major problems are selected to deal along with, which are service availability, contextualisation, and servitisation problems in the IoT paradigm. Considering these three problems and having an aspect of their solving in a transparent way, we map the state-of-the-art solver based on their integrated approaches, their mechanisms, and techniques in order to define the limitation based on each axis. Taking into count those findings, five contributions are proposed using the distributed algorithms and the autonomic computer to answer the research questions, the closure (safety) and the convergence (liveness) properties will be defined and proved satisfied by two proposed protocols. In addition, to highlight the importance of these protocols, we present an application in the context of IoT solutions.