SILICON PHOTONIC PHYSICAL UNCLONABLE FUNCTIONS by Brian Christopher Grubel A dissertation submitted to Johns Hopkins University in conformity with the requirements for the degree of Doctor of Philosophy. Baltimore, Maryland September 2017 © 2017 Brian Grubel All Rights Reserved Abstract Cryptographic systems control access to protected assets using keys stored in non-volatile digital memory which is vulnerable to tampering, substitution, and duplication. These key storage solutions require countermeasures that increase cost and complexity thus making their practical scalability to low-end systems intractable. Stored keys are usually derived from algorithms which include hash functions, random sequences, one-way permutations, and ciphers. Such approaches rely on computational asymmetry, e.g. algorithms that are easy to compute, yet difficult to reverse. However, the security of such “one-way” functions is not rigorously proven. Physical keys store private information within the physical structure of an object to protect against counterfeiting and spoofing. Recent physical key demonstrations are sufficiently complex to be believed unclonable; however, they exhibit reduced capacities for secret information, slow information generation rates, and vulnerability to emulation due to their linearity and slow speed. Here we investigate a primitive, a basic building block of a cryptographic system, based on the ultrafast and nonlinear optical interactions within an integrated silicon photonic cavity. Such a cavity can imprint a unique signature on an optical wave through the combination of a large number of spatial modes in a constrained medium. This interaction produces a highly complex and unpredictable, yet deterministic, ultrafast response that serves as a unique “fingerprint” of the cavity. This research includes methods to derive binary sequences from such waveforms that satisfy unpredictability, robustness, and entropic measures to meet security requirements for private key storage. We show that this robust physical key is unclonable, is impossible to emulate, and achieves dramatically improved capacity and throughput of secure information. We show that attempts at cloning this novel key fail, even with complete knowledge of the key design and maximally identical fabrication processes. We further demonstrate its use in authentication and communications ii applications and show that the device’s total information density is far beyond that of modern storage media. This innovative security measure is extremely small, inexpensive, and compatible with both conventional semiconductor integration and optical communications systems. This approach represents a revolutionary advancement in the construction of physical keys. Primary Reader: Dr. Amy C. Foster Secondary Reader: Dr. A. Brinton Cooper iii Acknowledgements There are many extraordinary people who have contributed to this project to help ideas become reality. I would like to thank my advisors: A. Brinton Cooper, Amy Foster, and Mark Foster for their example, advice, and encouragement. This multi-disciplinary work would not have been possible without their unique expertise and insight. As this work was unfunded at first and a deviation from their past research directions, I am eternally grateful for their willingness to venture into the unknown and provide both material and moral support. Further, if I have learned just one thing in my career it is that good administrative staff form some of the most important people in any organization. As such, I thank Debbie Race for her support in helping me navigate complex educational issues and maintain compliance with process and procedure. I appreciate Michael Kossey for the many weeks he spent at NIST building our prototype devices and teaching me vital laboratory practices. I have deep respect for Bryan Bosworth and the countless hours he spent in the laboratory with me troubleshooting experimental setups, engaging in philosophical conversation, and contemplating other distracting security projects all while ensuring that I did not break anything too important. I would also like to thank (and apologize to) Jasper Stroud, Milad Alemohammad, Jeff Shin, Kangmei Li, and Hongcheng Sun for their tolerance as I borrowed essential laboratory equipment at some of the most inconvenient times for them. I thank Arpad Neale and Dan Vresilovic for their boundless enthusiasm and their efforts on several side projects supporting this work. I thank Boeing for providing financial support as well as the support from my managers to include Glenn Hendershott, Naomi Hirose-Gormally, Tony Stamp, and Ann Stevens. I am also very appreciative of my colleagues Dave White, Dion Reid, and Izzy Rodriguez who originally introduced me to the field of information security and continued to provide me with their friendship iv and mentorship. I also extend my thanks to my former managers, mentors, and colleagues at Northrop Grumman who introduced me to the field of optics, especially Dick Leitholf, Carl Hikes, Jeff Peck, Tom Engel, and Kevin Hays. I thank Meghan for her support, love, kindness, and patience. She forced me to take the occasional break and keep things in perspective. I am beyond convinced that this achievement (and many others) would not have been possible without her. I thank my family and friends who have always supported me over the years and endured the awkward answers to their occasional questions about my research. And finally, I thank my parents, Arnold and Patricia, from whom I have learned countless lessons in addition to receiving their unwavering support, without which I would not have survived this program. They made many sacrifices such that I could get an education. I dedicate this work to them. v Nothing in this world can take the place of persistence. Talent will not: nothing is more common than unsuccessful men with talent. Genius will not; unrewarded genius is almost a proverb. Education will not: the world is full of educated derelicts. Persistence and determination alone are omnipotent. Calvin Coolidge 1872-1933 vi TABLE OF CONTENTS SECTION PAGE List of Tables ................................................................................................................................ xii List of Figures ............................................................................................................................... xii 1 Introduction ..........................................................................................................................1 1.1 Modern Security and Privacy Issues ................................................................................... 2 1.2 Physical Unclonable Functions ........................................................................................... 4 1.2.1 Concept .................................................................................................................................... 4 1.2.2 Properties and Component Requirements................................................................................ 8 1.2.3 Related Cryptographic Concepts ............................................................................................. 9 1.2.4 Previous Work ....................................................................................................................... 13 1.2.5 Testing and Specification of PUFs ........................................................................................ 21 1.3 Research Goals and Objectives ......................................................................................... 34 1.3.1 Integrated Silicon Photonics .................................................................................................. 34 1.3.2 Goals ...................................................................................................................................... 40 1.3.3 Objectives .............................................................................................................................. 41 1.4 Dissertation Organization ................................................................................................. 44 1.5 Summary ........................................................................................................................... 46 2 Light Transport through Chaotic Cavities .........................................................................48 2.1 Introduction ....................................................................................................................... 48 vii 2.2 Design ............................................................................................................................... 49 2.2.1 Chaos and Dynamical Billiards ............................................................................................. 49 2.2.2 Geometry ............................................................................................................................... 52 2.2.3 Design Concept ..................................................................................................................... 58 2.2.4 Design Summary ................................................................................................................... 61 2.3 Modeling and Simulation .................................................................................................. 62 2.3.1 Ray Tracing Model ................................................................................................................ 62 2.3.2 Waveguide