Projects per year
Abstract
This thesis unfolds a journey into the realm of cloud integrated systems. More specifically, it explores the transformational role of diverse cloud infrastructure, be it public or private, centralized or edge-based, when integrated into traditional systems. In this transformation, the cloud assumes the vital role of controllers. Inevitably, this shift towards cloud integration also brings into play the expansive network that acts as the connective tissue between traditional systems and the cloud, adding another layer of complexity to the newly formed integrated system.
In this work, we shed light on the less-talked-about side of cloud integration. Beyond the evident benefits of this transition, we face an array of challenges that emerge along with the introduction of the cloud and its accompanying network. Adapting traditional system deployment to this new era of cloud-based computing is one such necessity. The advent of virtualization and container technologies introduces additional requirements for software management. Shared infrastructure mandates stricter control over incoming traffic. Furthermore, real-world networks often act unpredictably, straying from their simulated behaviours. Even the much-touted 5G technology has not completely lived up to the expectations set a decade ago.
However, the ambition of this thesis does not lie in the enhancement of existing infrastructure, the improvement of cloud technologies, or the acceleration of network speed. Rather, it aims to accept and work within the limitations and flaws inherent in both cloud and network infrastructures. The primary goal is to recognize the chal- lenges these systems introduce, embrace their imperfections, and adapt our systems to work effectively with the realities of our imperfect cloud and unpredictable network environments.
To accomplish this, the thesis undertake a comprehensive analysis of two types of cloud integrated systems—Cloud RAN and Cloud Control System. A central focus is the evaluation of the practicality of implementing these systems using existing infrastructure. This evaluation is based on rigorous simulation as well as hands-on testbed experiments. In response to the insights gained from these assessments, the thesis proposes an innovative framework, built on a microservice architecture, to de- ploy cloud services more effectively for these systems. This framework is designed to mitigate the network latency impact brought on by unpredictable, “wild” environments. It does so by incorporating specialized prediction and estimation services, thereby enhancing the adaptability of these systems to real-world challenges.
In this work, we shed light on the less-talked-about side of cloud integration. Beyond the evident benefits of this transition, we face an array of challenges that emerge along with the introduction of the cloud and its accompanying network. Adapting traditional system deployment to this new era of cloud-based computing is one such necessity. The advent of virtualization and container technologies introduces additional requirements for software management. Shared infrastructure mandates stricter control over incoming traffic. Furthermore, real-world networks often act unpredictably, straying from their simulated behaviours. Even the much-touted 5G technology has not completely lived up to the expectations set a decade ago.
However, the ambition of this thesis does not lie in the enhancement of existing infrastructure, the improvement of cloud technologies, or the acceleration of network speed. Rather, it aims to accept and work within the limitations and flaws inherent in both cloud and network infrastructures. The primary goal is to recognize the chal- lenges these systems introduce, embrace their imperfections, and adapt our systems to work effectively with the realities of our imperfect cloud and unpredictable network environments.
To accomplish this, the thesis undertake a comprehensive analysis of two types of cloud integrated systems—Cloud RAN and Cloud Control System. A central focus is the evaluation of the practicality of implementing these systems using existing infrastructure. This evaluation is based on rigorous simulation as well as hands-on testbed experiments. In response to the insights gained from these assessments, the thesis proposes an innovative framework, built on a microservice architecture, to de- ploy cloud services more effectively for these systems. This framework is designed to mitigate the network latency impact brought on by unpredictable, “wild” environments. It does so by incorporating specialized prediction and estimation services, thereby enhancing the adaptability of these systems to real-world challenges.
Original language | English |
---|---|
Qualification | Doctor |
Supervisors/Advisors |
|
Award date | 2023 Dec 8 |
Publisher | |
ISBN (Print) | 978-91-8039-873-2 |
ISBN (electronic) | 978-91-8039-872-5 |
Publication status | Published - 2023 Nov 14 |
Bibliographical note
Defence detailsDate: 2023-12-08
Time: 09:15
Place: Lecture Hall E:1406, building E, Ole Römers väg 3, Faculty of Engineering LTH, Lund University, Lund. The dissertation will be live streamed, but part of the premises is to be excluded from the live stream.
External reviewer(s)
Name: Berger, Michael
Title: Assoc. Prof.
Affiliation: DTU Technical University of Denmark, Denmark.
---
Subject classification (UKÄ)
- Computer Systems
Fingerprint
Dive into the research topics of 'Taming Cloud Integrated Systems in the Wild'. Together they form a unique fingerprint.Projects
- 1 Finished
-
Design, optimisation and control of cloud-integrated autonomous networked systems
Peng, H. (Research student), Kihl, M. (Supervisor), Fitzgerald, E. (Assistant supervisor) & Tärneberg, W. (Assistant supervisor)
2017/11/13 → 2023/07/01
Project: Dissertation