Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Software Defined Networking for Survivable Optical Layer Provisioning and Network Security Using Machine Learning
In recent times the Internet has greatly enabled global communication which has led to an exponential increase in the number of users and worldwide traffic. The optical network plays a crucial role in meeting this demand and serves asthe Internet’s communication backbone. Software-Defined Optical Networking (SDON) has been proposed to support a large number of users at many different locations, with various different types of services and with diverse applications.SDN-based optical networking poses particular challenges but can also hold great potential. As the number of users and services grows, the optical network capacity has to grow accordingly and existing capacity needs to be utilized efficiently.In this dissertation, we comprehensively study the SDN paradigm in optical networks; in particular, our studies are focused on the data layer, the control layer, and solutions for efficient and survivable network service provisioning. Ourproposed solution provides excellent network performance, reducing the blocking probability of service requests, supporting service differentiation, and survivability for the whole network. We first present a novel control-layer mechanism called RDR (Resource Delayed Release). RDR combines an optimization method with SDN-enabled network management to improve network parameters and manage network resources actively. In the optical network, a fiber cut or a link failure can cause the loss of a significant amount of data. To overcome this type of failure, we evaluate different protection methods combined with the RDR method to reroute the traffic on a backup path to the destination. Next, we develop solutions for handling cyber attacks on the SDON network. We propose and evaluate a novel Hierarchical Classical Controller (HCC) algorithm to identify the attacks at an early stage to make the network more resilient while adhering to the controller’s capacity limitations. Then, we present an analysis of Internet 2 traffic data that can be used to gain insights into the network operation. Finally, we analyze the behavior of actual Internet Service Provider (ISP) traffic data. In particular, we perform data mining to identify patterns in order to predict congestion and the packet transfer rates.
Computer Engineering|Computer science
Mehr, Shideh Yavary, "Software Defined Networking for Survivable Optical Layer Provisioning and Network Security Using Machine Learning" (2022). ETD collection for University of Nebraska - Lincoln. AAI29322684.