VULNERABILITY ANALYSIS OF THE PHYSICAL AND LOGICAL NETWORK TOPOLOGY ON THE U.S. VIRGIN ISLANDS

Abstract

In 2017, two hurricanes, Irma and Maria, left the U.S. Virgin Islands with a destroyed telecommunications infrastructure, demolished homes and collapsed powerlines. Even though the communications system is broken into several sections (e.g., landline telephone, broadcast radio, and Internet service), the telecommunications network as a whole was severely impacted. Previous work has created a mapping and vulnerability analysis of the physical network infrastructure on the island of St. Croix, finding several single points of failure in the St. Croix network infrastructure. Data of the logical network infrastructure has been collected from the Center for Applied Internet Data Analysis (CAIDA) Ark Measurement Infrastructure, the Réseaux IP Européens (RIPE) Atlas Network, and the Naval Postgraduate School. This data is primarily traceroute data measuring the speed and route that messages take on their way to a specified destination. This thesis uses the traceroute data to create interface, router, and autonomous system-level network topologies of the U.S. Virgin Islands. We found that there are several nodes in the graph with high betweenness values, indicating that the network may be susceptible to congestion or disconnection during adverse events. To remedy this, we suggest adding redundancy to the important nodes or adding direct connections between distant nodes.