Common shared keys, which can be used for encryption/decryption, authentication and various other security primitives, plays a fundamental role in securing modern digital systems. Due to such significant importance, there has been a tremendous amount of work on the design of key generation/distribution schemes. Despite of these efforts, significant challenges, such as scalability and availability, remain to be fully addressed. This project investigates information theoretic approaches for secret key generation with focuses on fundamental and challenging issues under more realistic models. The resulting schemes are anticipated to be more easily applied to the design of practical wireless networks. The project will leverage collaboration with industries and will actively involve students with various backgrounds. In particular, the project will develop key generation schemes under the framework of key generation via public discussion. The standard scheme under this framework is a so-called "omniscience" scheme. However, such a scheme requires all nodes in the network to transmit, assumes the public discussion is heard noiselessly by all nodes in the network, and does not take the implementation complexity into consideration. While these assumptions enable the tractability of the problems and facilitate to obtain crucial insights, in order for such studies to provide useful and accurate guidance in practice, realistic constraints must be taken into consideration, which pose three fundamental directions that this project will explore. In this project, thrust 1 focuses on key generation with vocality constraints that addresses the following fundamental question: for a general source network, is it possible to design schemes that involve only a subset of nodes to talk but still achieve the key capacity? Thrust 2 focuses on key generation with physical channel constraints. In particular, motivated by constraints associated with physical channels, the project will design key generation scheme for scenarios when the public discussion of each node can only be received by nodes in its transmission range and the simultaneous transmission of multiple nodes will interfere with each other. Thrust 3 focuses on key generation with complexity constraints and investigates how to minimize the complexity in the design of key generation schemes and how to reduce computational resource consumption required for key generation schemes.
CIF: Small: Collaborative Research: Secret Key Generation Under Resource Constraints