Mathematics and statistics

A wide range of cloud services and applications operate on sensitive data such as business, personal, and governmental information. This renders security and privacy as the most critical concerns in the cloud era. The objective of this project is to question the separation approach in the design of security and reliability features of storage systems, and to investigate new, coding-based security mechanisms based on a joint-design principle. The proposed program will result in a myriad of outcomes.

Side-channel attack (SCA) has shown to be a serious implementation attack to many cryptosystems. Practical countermeasures only mitigate the vulnerability to some extent. Considerable research efforts on leakage-resilient cryptography have so far not led to practical leakage-resilient implementations. One hindering reason is the lack of commonly accepted and sound metrics, standards, and evaluation procedures to measure and evaluate the vulnerability/resilience of cryptosystems to various side-channel attacks.

The project aims at quantifying a general network's inner potential for supporting various forms of security by achieving secret common randomness between pairs or groups of its nodes. Statistical and computational secrecy measures are being considered against a general passive adversary. Common-randomness-achieving protocols are classified into two groups: culture-building and crowd-shielding. The former achieves common randomness between nodes situated in close proximity of each other, from correlated observations of specific (natural or induced) network phenomena.

Information sharing between operators (agents) in critical infrastructure systems such as the Smart Grid is fundamental to reliable and sustained operation. The contention, however, between sharing data for system stability and reliability (utility) and withholding data for competitive advantage (privacy) has stymied data sharing in such systems, sometimes with catastrophic consequences. This motivates a data sharing framework that addresses the competitive interests and information leakage concerns of agents and enables timely and controlled information exchange.

The need for accurate and unforgeable identity recognition techniques has become an issue of increasing urgency. Biometric approaches such as iris recognition hold huge promise but still have significant limitations, including susceptibility to 'spoofing'. This project seeks to advance our knowledge of security and accuracy of multibiometric systems by inventing, evaluating, and applying innovative methods and tools to combine highly accurate static traits, such as iris patterns, with novel traits based on the dynamics of eye movements.