A comprehensive threat assessment framework for securing emerging technologies
Sanzgiri, Ameya M.
MetadataShow full item record
Wireless devices are becoming an integral part of the human environment and their seamless integration has created a range of new wireless sensor network architectures. Unfortunately the security of such networks often lags behind other advances and more often than not is developed only after the core systems and protocols have been standardized. This results in these security schemes having case-specic reactive attributes and being unable to anticipate any changes in the attacker's attack vector. Integrating security into the next generation computer applications' core design is paramount as traditional "reactive" security operations on top of normal functionality will be an expensive and ineffective proposition. The primary focus of this dissertation is to develop a framework that assists in the formulation of proactive security schemes. A proactive security scheme aims at dissuading an adversary from attacking a system by increasing the cost of attack. Such schemes need to be integral to the design of the emerging technologies' domain, so that protection against attacks, especially the stealthy and smart ones can be devised. However, to effectively design such schemes, one needs to understand the threats to a system as well as their effects on a system. Threat modeling in itself is a significant research challenge due to the lack of easy to understand techniques or methodologies. This dissertation focusses on the problem of creating a framework which is generic enough for emerging systems and networked applications and can be easily adapted to provide a defender with appropriate attack vectors and risk analysis capabilities. First the dissertation presents a paradigm shift in threat modeling by incorporating the attacker's perspective in the implementation of an attack and analyzing the various factors that an attacker would have to consider in his attack. Second, the identification of the avenues where the proposed framework can be used to increase the effectiveness of the modeling techniques is discussed. Although the framework can be used at any abstracted level, the dissertation focusses on some of the most important avenues of attacks by studying the problem of identifying levels which present the most likelihood of risks. The levels discussed are Architecture level - where the model is applied to the entire architecture, considering the specifics of the architecture and investigating threats to the architecture; Protocol level - where protocol (network) specifics and threats to the protocols are considered and finally, the Application level - where the threat model considers the application specifics, such as the purpose of the application and the unique features of the applications as well as the information from the architectural and protocol level threat modeling. The framework is applied to several existing as well as emerging real-world applications and open-ended attacks to identify and analyze the risks faced by these applications by using epidemic theory, probabilistic modeling and concepts from complexity theory to conduct the risk analysis. The risk verification is done via simulations as well as real world experiments. The aim of this research is to develop a framework, which will be a valuable aid in the creation of sound security schemes and risk analysis in the future.