## ELE 538: Information Theoretic SecurityProf. Paul Cuff, Princeton University, Fall Semester 2013-14.
## Course DescriptionThis course focuses on secure communication built on information theory, which does not assume that an adversary has computational limitations. We begin with classic results on secrecy systems by Shannon and Wyner, paving the way for a study of modern research on physical-layer security, secret key agreement, etc. We consider distortion guarantee as a metric for partial secrecy and compare to the more traditional use of equivocation. The course also touches on other aspects of secure communication, such as authentication. Related topics such as digital watermarking and information embedding will also be discussed. Prerequisites: Exposure to information theory or concurrent enrollment in ELE 528. ## TopicsApplied cipher systems: Vernam cipher and one-time-pad, modern cryptography, Diffie-Hellman, RSA. AES and side-channel attacks. Authentication. Information theory foundation: Shannon cipher analysis. Wyner's wiretap channel. Gaussian channels (with and without channel state). Multiuser settings. Strong vs. weak secrecy. Rate-distortion theory for secrecy systems: distributed channel synthesis; alternative proofs based on cloud mixing, random binning, and method of types. Equivocation as a special case under log-loss distortion. Secret key agreement: common information; channel reciprocity (fading); quantum key exchange; privacy amplification. Other topics: coding with linear codes; authentication; watermarking. ## Teaching Staff## InstructorProf. Paul Cuff ## Time and Location## LecturesRoom: B-205 E-Quad ## TextNo textbook purchase is required. The following textbook is a useful general resource for information theory: Elements of Information Theory, second edition, Cover and Thomas. The next two textbooks are good research references, covering a wide variety of relevant results. Both cover much of the same content, although the first is obviously more security focused. Physical-Layer Security, Bloch and Barros. Network Information Theory, El Gamal and Kim. ## BlackboardBlackboard will not be used for this course. |