A talk by: Dr. George Loukas – University of Greenwich
ABSTRACT: In the past, the primary aim of attacks in cyber space would be to cause damage in cyber space. This is not any more the case. We have become so dependent on computer communications, social networks, Internet of Things devices and networked cyber-physical systems that cyber attacks can aim to damage our physical belongings, our health, our emotions, our vehicles and even our critical infrastructures. This has elevated the importance but also the challenge of detecting cyber threats. In this talk, we discuss two new approaches for detecting threats in challenging cyber environments. The first approach is on detecting attacks against cyber-physical systems, with a case study on a four-wheel drive robotic vehicle. Here, we show that the physical manifestation of different cyber attacks on a cyber-physical system is not only a challenge, but also an opportunity in helping detect them more rapidly and more accurately than if only traditional cyber monitoring were used. The second approach is on detecting attacks that employ deception (e.g., spear-phishing, obfuscated URLs, drive-by downloads, spoofed websites, scareware etc.) to circumvent traditional technical security controls and target the human user directly. Here, we employ a Human as a Security Sensor (HaaSS) approach, where it is information from the users than from technical security systems that helps detect an attack. For such an approach to be practical, we need to be able to predict how good a user can be at sensing cyber threats, ethically, automatically and in real-time. In terms of technologies, the two approaches described in this talk use a variety of techniques from the fields of statistical machine learning, deep learning, network modelling and computation offloading.
Dr. George Loukas is a Senior Lecturer in cyber security at the University of Greenwich, U.K. He is currently the Principal Investigator for three large-scale EU and UK research projects, focusing on the security of autonomous vehicles, securing collaboration of communities and law enforcement agencies, and bridging emotion research with cyber security in the context of smart home environments. His “Cyber-physical attacks: A growing invisible threat” book has received highly positive full-length reviews by the IEEE Cipher magazine, the Professional Security Magazine, and the Policing Journal, and has been chosen by ACM in the top 10 notable items in the Computing Milieux category published in computing in 2015.
For more information, please visit here: http://www.georgeloukas.com/
The slides of the presentation can be accessed here.