Open Access Repository

Reliable scheduling and resource allocation for IoT applications in fog computing

Naha, RK ORCID: 0000-0003-4165-9349 2021 , 'Reliable scheduling and resource allocation for IoT applications in fog computing', PhD thesis, University of Tasmania.

Full text not available from this repository.

Abstract

The massive development of ubiquitous computing leads the modern world to latency sensitive Internet of Things (IoT) applications such as smart transportation systems, healthcare services and emergency response services, in order to enable a better quality of life. These services generate a huge amount of data which need to be processed near to the edge. These kinds of applications can be handled effectively and efficiently using Cloud infrastructure because of the on-demand services and scalability feature of the Cloud. However, managing IoT applications in the Cloud exclusively is not a good solution for some applications, especially for those which are latency-sensitive. Thus, Fog computing has emerged which resides between the Cloud and the end devices. Generally, IoT devices and sensors are connected to the Fog devices. These Fog devices are located in close proximity and are responsible for intermediate computation and storage.
Allocating resources for user applications to the highly dynamic, heterogeneous and complex Fog environment is challenging. In addition, the user might change their requirements dynamically and also require better reliability from the providers. Hence, reliable resource allocation and task scheduling techniques are required which consider the dynamic behaviour of the users’ requirements. In Fog devices, available resources are changing over time since they are not dedicated to running Fog applications. More-over, resource failure and link failure could occur frequently in the Fog environment because of a lack of central management, the autonomous characteristics of the devices and wireless connectivity. Furthermore, most of the Fog devices are battery-powered; therefore, resource allocation and scheduling techniques need to be energy-efficient. To address these problems, this thesis proposes several resource allocation and failure handling techniques to make the Fog environment reliable. The influence of strict execution time and data transfer time is also considered during resource allocation and scheduling. In the Fog, some users may request cost-effectiveness, rather than fast execution. Hence, cost-effectiveness is also investigated. An evaluation of the proposed methods was tested in a simulated environment.
This thesis adds to the body of the knowledge by making the following contributions:
1. An extensive survey on architecture, resource allocation and scheduling and failure handling in the Fog computing environment.
2. A comprehensive study on Fog computing architecture to develop a simulation environment for Fog computing.
3. A deadline-based dynamic resource allocation and provisioning in an hierarchical and hybrid fashion with dynamic user behaviour.
4. A multi-criteria-based dynamic user behaviour aware resource allocation in which resources are dynamic.
5. A fuzzy logic-based failure handling mechanism to handle predicted and unpredicted failures in the Fog environment, in order to ensure robust scheduling.
6. A multiple linear regression-based energy-aware resource allocation mechanism to ensure reliable execution when most of the devices have limited available energy for operation.

Item Type: Thesis - PhD
Authors/Creators:Naha, RK
Keywords: Fog computing, Internet of Things (IoT), Resource allocation, Resource provisioning, Time-sensitive application, Dynamic behaviour, Application scheduling
Copyright Information:

Copyright 2021 the author

Item Statistics: View statistics for this item

Actions (login required)

Item Control Page Item Control Page
TOP