Projects
The focus of this project is to achieve secure communication within a network by using covert channels. Security is achieved by hiding node endpoints (IP and MAC addresses) and communication (payload data) from inside attackers and other adversaries. Covert messages are blended into suitably varying active network traffic. Throughput depends on detectability and reliability requirements and scales based on the size and characteristics of the network.
My work focuses on identifying sequences of semantically rich log events that are shared among malware instances.
These events can be labeled by analysts to reduce duplicate work and to facilitate analysis of
multi-purpose malware. These labels will also be used to build
a automatic behavior-based malware classification systems.
Survivability analysis focuses on the ability of network entities to function during incidents such as attacks.
Currently, testing survivability of ad hoc networks consists of
running scenarios with several configurations, often thousands,
to obtain an understanding of the impacts of an attack. This
process is very latent, choice of configurations are subjective or
random, and results do not generalize to different scenarios.
My work focuses on alleviating these problems by researching
novel methods for efficient survivability analysis leveraging machine
learning and novel attacker-focused network representations.
Previous work in automated penetration testing, in general, utilizes attack graphs based on known software vulnerabilities. This work focuses on developing higher fidelity penetration test plans that incorporate the information gathering phase (including network sniffing, intelligence gathering, etc.) in both infrastructure and mobile ad-hoc networks