Quantifying Urban Road Network Vulnerability and Resilience to Attacks

TL;DR

This paper presents a framework combining complex network theory and real-time data to assess and improve the resilience of urban road networks against cyber-attacks, demonstrated on Boston’s transportation system.

Contribution

It introduces a novel, integrated approach for quantifying and detecting vulnerabilities in city-scale transportation networks using complex network analysis and real-time data streams.

Findings

Targeted attacks on few nodes cause large disruptions.

Real-time data can predict and mitigate cyber-attack impacts.

Framework applicable to other urban transportation systems.

Abstract

The rise of connected and autonomous vehicles, combined with the proliferation of IoT and connected surfaces, lead to the emergence of novel complex cyber risks. Lack of encryption and authentication in internal vehicular networks are widely recognized as cause for concern by cybersecurity experts, automobile, and OEM manufacturers. This concern has only been growing with the increase in cybersecurity incidents and demonstrations showing different vehicular vulnerabilities, making it nearly impossible to completely secure vehicles against cyber-attacks. Of particular concern is the potential for large-scale vehicular cyber-attacks to cascade to transportation networks, which are the lifeline of cities. Here, we develop a framework based on complex network theory, traffic flow, and new data based technologies to quantify the vulnerability of city-scale transportation to cyber-attacks. Application of our framework to the road network of Boston reveals that targeted attacks on a small fraction of nodes leads to disproportionately larger disruptions of routes. We develop an early-detection framework to quantify real-time risk based on gathering multidimensional traffic flow, incident, and social media data sets. Our results illustrate an effects based approach to transportation cyber-defense, through informed, intelligent vehicular agents.