Author ORCID Identifier

0000-0002-1056-880X

Date of Award

12-12-2022

Degree Type

Dissertation

Degree Name

Doctor of Philosophy in Business Administration (PhD)

Department

Computer Science

First Advisor

Xiaojun Cao

Second Advisor

Anu Bourgeois

Third Advisor

Yanqing Zhang

Fourth Advisor

Yichuan Zhao

Abstract

In the next generation of networking paradigms, network services are expected to be pro- vided with low latency, high reliability, and high connection density. With network function virtualization (NFV), these network services can be delivered via software-based modules known as service functions. While NFV enables greater flexibility and efficiency in delivering service functions, it poses challenges for network robustness and reliability. In order to maintain network reliability and prevent service interruptions caused by network failures, it is essential to allocate backup resources properly. In this dissertation, we investigate service function protection in NFV and develop efficient methods for allocating backup resources.

First, we discuss how to protect NFV services from service function instance failures in unique function networks. Second, we investigate how to ensure that NFV services are effectively protected from service function instance failure in multiple function networks. Third, we investigate how to adequately protect NFV services from service function forwarder failure in cloud networks. Our fourth study explores how to protect NFV services against service function forwarder failure in edge networks. The fifth part presents an efficient method for protecting NFV services against service function chain failures in the edge networks. We analyze and discuss our proposed protection mechanisms and conduct extensive simulations to assess their effectiveness. Simulation results demonstrate that the proposed protection mechanisms are effective and efficient under various conditions. Lastly, we suggest two possible research directions for future investigation.

File Upload Confirmation

1

Available for download on Sunday, October 20, 2024

Share

COinS