Pattern Analysis of Anomalous Behavior of a Heterogeneous Software Defined Wireline Public Telecommunications Network Before and During the COVID-19 Pandemic

Paxton J Louis, Pace University

Abstract

This dissertation analyzed the traffic data from a real-world wireline public telecommunications network to identify patterns of anomalous network and user behavior with a focus on developing a customizable and scalable methodology to analyze telecommunications traffic data using open-source machine learning tools that can be customized for specific big data analytical needs. The focus of my research is a wireline public telecommunications carrier network that is a software-defined Internet Protocol based network, which supports a wide variety of customer types. My research led me to identify data patterns (fingerprints) in the operations of a real-world operating wireline public telecommunications network. The patterns found can be used as the basis for new machine learning and AI-based pattern recognition algorithms, predictive algorithms, or design methodologies to improve network designs and operations. The data I had collected before the pandemic served as my baseline for customers’ and the network’s normal pre-COVID-19 pandemic behaviors. The data was not scrubbed by the telecommunications carrier. My collection of data before and during the early days of the COVID-19 pandemic was purely coincidental. However, as tragic as the Pandemic was, I saw the opportunity to record data during an extraordinary period of time.In my 40 Plus Years of experience as a telecommunications network engineer, network signaling subject matter expert, and network manager, I have had the opportunity to design, manage, and operate telecommunications carrier networks during a time when using a slide rule was the norm, the electronic portable calculator was a new device and the desktop computer did not exist. Early in my career, automated pattern visualization was not used by network managers or network engineers, rather we had to read tables and lists that were manually created. The customizable and scalable methodology I developed and the software tools I identified, enabled me to demonstrate that a data scientist could analyze big data using a laptop and cloud-based computing resources and proved my hypotheses.

Subject Area

Computer science|Computer Engineering|Engineering|Artificial intelligence

Recommended Citation

Louis, Paxton J, "Pattern Analysis of Anomalous Behavior of a Heterogeneous Software Defined Wireline Public Telecommunications Network Before and During the COVID-19 Pandemic" (2023). ETD Collection for Pace University. AAI30638740.
https://digitalcommons.pace.edu/dissertations/AAI30638740

COinS