Graph-Theoretic Analysis and Design of Communication Networks
Department
Electrical and Computer Engineering
Major
Computer Engineering and Electrical Engineering
Research Advisor
Çetinkaya, Egemen K.
Advisor's Department
Electrical and Computer Engineering
Abstract
An important aspect of the design of communication networks is the resilience of the network. A resilient network is one that can withstand distributed damage from a targeted attack or random failures while maintaining a majority of its underlying connection structure (i.e., not becoming disconnected). The main goal of this project is to conduct a graph-theoretic analysis of networks in order to gain a deeper understanding and optimize networks against potential attacks or environmental damages. The project utilizes graph theory and a variety of software tools, particularly Python and the NetworkX software packages, with which we can simulate the network structure and analyze it based on a variety of metrics. We will focus on the spectra of graphs as the graph metric, which are the eigenvalues and multiplicities of the normalized Laplacian matrix. Graph spectrum is a useful metric to study the internal structure of complex networks and how they react when components are removed or added since it is especially helpful in our understanding and visualization of networks containing hundreds of nodes and links. Using metrics such as this, we can also study a variety of network types besides communication networks, such as social networks (Twitter metadata, Facebook friends, etc.) or infrastructure networks (power grids, roadways, etc.). Analyzing these networks, we can gain a deeper understanding of how complex networks are structured and use our findings for a variety of beneficial applications.
Biography
Tristan Shatto is a 2nd semester junior that is dual majoring in Computer Engineering and Electrical Engineering at Missouri S&T. Originally from the Kansas City area, he has been working as an undergraduate research assistant under Dr. Egemen Çetinkaya of the Electrical and Computer Engineering department for the past 2 years; His main area of research is the graph-theoretic analysis of complex networks. He is a member of IEEE, and has had his paper entitled “Spectral Analysis of Backbone Networks Against Targeted Attacks” published in the International Conference on the Design of Reliable Communication Networks (DRCN) 2017. Tristan wishes to pursue a career in the aerospace and defense industry after graduation.
Presentation Type
OURE Fellows Proposal Oral Applicant
Document Type
Presentation
Location
Turner Room
Presentation Date
11 Apr 2017, 2:20 pm - 2:40 pm
Graph-Theoretic Analysis and Design of Communication Networks
Turner Room
An important aspect of the design of communication networks is the resilience of the network. A resilient network is one that can withstand distributed damage from a targeted attack or random failures while maintaining a majority of its underlying connection structure (i.e., not becoming disconnected). The main goal of this project is to conduct a graph-theoretic analysis of networks in order to gain a deeper understanding and optimize networks against potential attacks or environmental damages. The project utilizes graph theory and a variety of software tools, particularly Python and the NetworkX software packages, with which we can simulate the network structure and analyze it based on a variety of metrics. We will focus on the spectra of graphs as the graph metric, which are the eigenvalues and multiplicities of the normalized Laplacian matrix. Graph spectrum is a useful metric to study the internal structure of complex networks and how they react when components are removed or added since it is especially helpful in our understanding and visualization of networks containing hundreds of nodes and links. Using metrics such as this, we can also study a variety of network types besides communication networks, such as social networks (Twitter metadata, Facebook friends, etc.) or infrastructure networks (power grids, roadways, etc.). Analyzing these networks, we can gain a deeper understanding of how complex networks are structured and use our findings for a variety of beneficial applications.
