Modelling and placement optimization of compound services in a converged infrastructure of cloud computing and internet of things

Abstract : The convergence of the Internet of Things (IoT) and Cloud Computing technologies is a promising approach. On the one hand, Cloud Computing provides virtually unlimited computing, networking, and storage resources for constrained IoT devices. On the other hand, the IoT enables the interaction of cloud services with real world things. Such integration stimulates innovation in both areas and provides novel service delivery models such as the Sensing as a Service in different application domains (i.e. healthcare, transportation, smart-city, smartbuilding). In such convergence, things are abstracted and offered as cloud services accessible over the Internet from any place and at any time. Such abstractions are known as Virtual Objects (VOs) and connect things to traditional cloud services (e.g. data analytics, storage services) to deliver IoT applications. In this thesis, we consider first a partial integration of the IoT and Cloud Computing. Such integration focuses on delivering the IoT within a single service level of Cloud Computing, namely: the application, the platform, or the infrastructure level. In this context, IoT and Cloud Computing resources are provisioned separately. We focus in this work on the orchestration of VOs in a cloud infrastructure. For this purpose, we define a provisioning algorithm based on a sharing strategy where each connected object is associated with a single VO and can be consumed by multiple applications. We propose two linear programs to perform the provisioning of VOs. The first considers no previously deployed VOs in the infrastructure, while the other takes into consideration pre-deployed VOs. Our approach minimizes VOs operational cost and communication latency in both cases compared to those with a non-sharing strategy. The second part of this thesis considers a full integration of the IoT and Cloud Computing. We refer to such integration as the Cloud of Things (CoT). In this context, a customer should be able to request end-to-end IoT application provisioning, deployment, auto-scaling, and release on the fly with minimal management efforts. In this thesis, we address the provisioning aspect. We define a resource-oriented model able to describe an IoT application request and a CoT infrastructure on different service levels. We base our model on the OCCI specifications defined by the OGF. Furthermore, we define a single stage provisioning algorithm to orchestrate a described IoT application into a CoT infrastructure. The algorithm considers cloud and IoT resources simultaneously. Simulations show that a one-stage provisioning process is 10% − 20% more efficient than two separate orchestration processes for cloud and IoT resources.
Complete list of metadatas

Cited literature [143 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/tel-01761673
Contributor : Frédéric Davesne <>
Submitted on : Monday, April 9, 2018 - 12:23:46 PM
Last modification on : Monday, October 28, 2019 - 10:50:22 AM

File

2017SACLE030.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : tel-01761673, version 1

Citation

Elie Rachkidi. Modelling and placement optimization of compound services in a converged infrastructure of cloud computing and internet of things. Networking and Internet Architecture [cs.NI]. Université Paris-Saclay; Université d'Evry-Val-d'Essonne, 2017. English. ⟨NNT : 2017SACLE030⟩. ⟨tel-01761673⟩

Share

Metrics

Record views

193

Files downloads

676