Activity
Time:
Ph.D. Dissertation Defense: Shideh Yavary Mehr
Date:
3:30 pm –
4:30 pm
Zoom Room: https://unl.zoom.us/j/92750914158
“Software Defined Networking for Survivable Optical Layer Provisioning, Network Security Using Machine Learning and an Intelligent Traffic Engineering Method”
In recent times, the internet has greatly enabled global traffic and led to an exponential increase in the number of users worldwide. Optical networks play a crucial role in meeting stringent real-time traffic demands and serves as the Internet’s primary communication backbone. Software-Defined Optical Networking (SDON) has been proposed to support a large number of users at many different locations due to its ability to support various types of services and applications. Despite its potential, SDN-based optical networking poses particular challenges due to the divergent real-time requirements that come with the diversity of applications, users, software and hardware. This thesis addresses three major challenges. (1) It provides a solution for efficient management of the control-layer by proposing a survivable network service provisioning and a protection mechanism to address a case of fiber cut or link failure that cause a loss of significant amount of data. (2) A machine learning-based solution for handling cyber-attacks on the SDON network using well-known classification techniques tailored to identify malicious traffic in real-time and thereby prevent a network failure. (3) We present a traffic engineering mechanism utilizing two types of extensive network traces, namely, Research and Education Network called Internet2 and a local commercial ISP operating out of the state of Nebraska. Internet2 (I2) traffic provides a new exhilarating challenge because it exists on an intricate, fast-growing network connecting a unique set of users. Optical fiber serves as the primary backbone of Internet2 and hence such networks need accurate analysis and measurement of the Internet traffic to supply the excellent services that users expect and predict future demands.
Committee Members:
Professor Byrav Ramamurthy, Advisor
Professor Massimiliano Pierobon
Professor Lisong Xu
Professor Fred Choobineh
In recent times, the internet has greatly enabled global traffic and led to an exponential increase in the number of users worldwide. Optical networks play a crucial role in meeting stringent real-time traffic demands and serves as the Internet’s primary communication backbone. Software-Defined Optical Networking (SDON) has been proposed to support a large number of users at many different locations due to its ability to support various types of services and applications. Despite its potential, SDN-based optical networking poses particular challenges due to the divergent real-time requirements that come with the diversity of applications, users, software and hardware. This thesis addresses three major challenges. (1) It provides a solution for efficient management of the control-layer by proposing a survivable network service provisioning and a protection mechanism to address a case of fiber cut or link failure that cause a loss of significant amount of data. (2) A machine learning-based solution for handling cyber-attacks on the SDON network using well-known classification techniques tailored to identify malicious traffic in real-time and thereby prevent a network failure. (3) We present a traffic engineering mechanism utilizing two types of extensive network traces, namely, Research and Education Network called Internet2 and a local commercial ISP operating out of the state of Nebraska. Internet2 (I2) traffic provides a new exhilarating challenge because it exists on an intricate, fast-growing network connecting a unique set of users. Optical fiber serves as the primary backbone of Internet2 and hence such networks need accurate analysis and measurement of the Internet traffic to supply the excellent services that users expect and predict future demands.
Committee Members:
Professor Byrav Ramamurthy, Advisor
Professor Massimiliano Pierobon
Professor Lisong Xu
Professor Fred Choobineh