Grant: Discovery Grants Program - Individual

The proposed research program is comprised of the following two research projects.1. Fault Tolerant CryptographyMost cryptographic algorithms and protocols involve massive computations and communications. Errors may occur from time to time. When there is an error, for an example, in the digital signature verification process, it is hard to tell whether this is a fake digital signature or there is a computation and/or communication error. The algorithm based fault tolerant (ABFT) approach achieves fault tolerance by tailoring the fault tolerance scheme to the algorithm to be performed. ABFT approach requires less overhead, uses the same type of computation as the targeted algorithm and the throughput will not be affected until an error is detected. The limitation of the ABFT approach is that it only applies to certain types of algorithms. In this research we develop an extended ABFS (EABFT) approach that can be applied to all well known cryptographic algorithms, including stream ciphers, AES block cipher, RSA and ECC public-key algorithms. In addition, we will design and implement a byte-wise 2-D fault tolerant AES systolic array for applications where high speed and high reliability are required. 2. Advanced Multilayer Access Control ModelingA document in a digital library could contain different types of content with different levels of privacy and/or restrictive information. A user should be able to access certain parts of the document according to his/her access rights and needs. In this project, we are going to develop an advanced multilayer access control model such that access to different types of content and different layers of security information can be done efficiently and effectively. The proposed advanced multilayer access control model can be integrated with other well known access control models, including role-based, mandatory and discretional access control models. It can also be implemented as a part of an operating system. The applications of the proposed access control model include Web based multimedia applications, digital libraries, e-education and e-health. In addition, we will investigate a method to convert the access control right rules for an application into a set of security criteria. The set of security criteria will be used as the inputs of the system design. As a result, it will significantly reduce the cost during system design and implementation due to any changes to the access control rules.