With the goal of harnessing the untapped potential of Iranian-Americans, and to build the capacity of the Iranian diaspora in effecting positive change in the U.S. and around the world, the Iranian Americans’ Contributions Project (IACP) has launched a series of interviews that explore the personal and professional backgrounds of prominent Iranian-Americans who have made seminal contributions to their fields of endeavour. We examine lives and journeys that have led to significant achievements in the worlds of science, technology, finance, medicine, law, the arts and numerous other endeavors. Our latest interviewee is Farinaz Koushanfar.
Farinaz Koushanfar is a Professor and Henry Booker Faculty Scholar in the Department of Electrical and Computer Engineering at the University of California San Diego. She is the co-founder and co-director of Machine-Integrated Computing and Security (MICS), an upcoming UC San Diego engineering research center which focuses on technical innovation and women empowerment. She is a fellow of the Kavli frontiers of engineering of the National Academy of Engineering, and was selected to the world’s top 35 innovators under 35 (TR-35). She serves as an associate partner of the Intel Collaborative Research Institute for Secure Computing. She has received a number of awards and honors for her research, mentorship and teaching, including the Presidential Early Career Award for Scientists and Engineers from President Obama, the ACM SIGDA Outstanding New Faculty Award, Cisco IoT Security Grand Challenge Award, as well as Young Faculty/CAREER Awards from NSF, DARPA, ONR and ARO. She received her Ph.D. from the University of California Berkeley.
Tell our readers where you grew up and walk us through your background. How did your family and surroundings influence you in your formative years?
My family emphasized and valued education. They instilled confidence in me by believing in my abilities from an early age and giving me opportunities to grow. For example, I was passionate about colors and shapes and have been obsessed with them since I was about 2 years old. I would spend hours matching them, drawing and painting. My father had a patient who held painting classes for older kids starting at 6-7+ years old. My parents had the art school assess me, and they made an exception so that I could start taking lessons even at 3.5 years old. Like many other kids in my generation and beyond, I was exposed to several extracurricular art and sports activities through the years. However, it was clear from my parent’s value system that mathematics and science were the priorities.
I also loved books and reading. We spent a lot of time at my grandparents’ house during my formative years. My grandfather had a big library filled with various Farsi, French, and English books that prompted me to learn other alphabets in addition to Farsi's in early age. He also had a few collections of 3D moulage kits that soon became my LEGOs. I played with them so much that he let me take them home. I was more interested in the shapes and putting the patterns together than in the biology of the kits. My grandfather, who studied in France and practiced medicine in the US after World War II, considered Marie Curie as the ultimate woman scientist and was impressed with the progress in nuclear physics. He often encouraged me to consider getting a doctorate in nuclear physics.
You received a number of awards and honors for your research, mentorship and teaching. What were the significant accomplishments that led to these?
The work in my lab is focused on bridging the gap between computer engineering and two critical areas in contemporary computing, namely security and machine learning. In terms of security, our work has pioneered a new genre of hardware security and IP protection techniques based on active post-fabrication control of silicon chips, and established new roots for hardware-based protection and trust. We invented methods that, for the first time, can uniquely track chips post-fabrication and in-field, or can detect hardware malware. Our hardware-based methodologies also significantly contribute to privacy-preserving computing.
In terms of machine learning, our projects are focused on creating disruptive automated methods, algorithms, and systems that serve the present demands of constrained embedded systems and enable novel types of real-time data-intensive applications. In recent years, machine learning and data-analytic experts have made impressive progress in adapting algorithms to the geometric structure of the data. However, concerns over scalability and ease of use present roadblocks to the efficient adoption of these algorithms on constrained platforms. Real life examples include drones or robots that have to run analytics and make real-time decisions with a limited energy supply. Our solutions target this space. As another example, our unique machine-integrated analytics approach helped us to prototype and report the first-ever training of complex deep learning networks (a widely popular contemporary machine learning algorithm) on a constrained mobile platform.
At UC San Diego, we are taking these ideas to the next level. Together with a colleague and collaborator, Professor Tara Javidi, we are establishing a new research center called Machine-Integrated Computing and Security (MICS), which we anticipate to launch this fall. The goal of this center is to develop a holistic approach to computing and security. We envision that this new approach could replace existing computing methods that involve collecting massive amounts of data blindly, learning data and models oblivious to the machine, or addressing security and hardware issues post-facto.
What has been your personal key to success? What were the biggest inspirations for your career?
The major inspirations for my career are my love for science and engineering, my curiosity to learn, my quest to make a difference, and the joy of working with others while educating the next generation of engineers and leaders.
Success is a relative term. I try to set near and long-term goals for my research and ventures, and plan my way towards achieving those milestones. My path often gets twisted or shifted and doesn’t exactly lead to where I intended to go, but I still find that the hypothetical goals help me to be inspired and focused. Beyond my family, my career has been significantly influenced by the advice, mentorship, and supervision of several excellent peers, colleagues, mentors, teachers, and academics.
As a researcher, faculty member, and educator, the most significant accomplishment for me has been empowering others — not just my own students and advisees, but also other students, researchers, and practitioners who are leveraging the results produced by my group or are otherwise influenced by our work. One particular topic that I have been passionate about for many years is empowering women.
Your research contributions include the creation of TinyGarble, the first scalable sequential methodology and libraries for implementation and optimization of the classic problem of secure function evaluation. Could you please shed light on this work? In particular, what helps you in creation of such new solutions?
With my hat on as a computer engineer, my main focus is on providing scalable and practicable computing solutions. Several of my research contributions are about closing the gaps between content, algorithm, and hardware, such that intensive or impractical applications can be finally scaled up and adopted. I recognized early in my career that a computer engineering background alone might be insufficient to address the many problems that interest me. Thus, while pursuing a Ph.D. in electrical engineering and computer science at UC Berkeley, I took several extra courses and ended up getting an additional degree in statistics and machine learning. Acquiring the broader knowledge has been tremendously beneficial for my career, allowing me to make connections between seemingly disparate but indeed relevant problems.
In the field of secure function evaluation, the main problem researchers face is how parties can compute a joint function while each party keeps its data private. The famous example is the problem where two millionaires compare their wealth without showing to each other (or to other parties) their wealth values. Our focus has been on a popular classic secure evaluation problem, called garbled circuit protocol, which was introduced by the famous computer scientist Andrew Yao more than thirty years ago. Despite significant progress throughout the past 30+ years, the existing solutions were still inefficient and did not scale well for more complex functions.
Our solution, called TinyGarble, altered the implementation landscape for this important problem by building upon and adapting concepts and tools from logic design — the same ones that facilitated automated scaling of contemporary silicon processors to billions of gates. Experiments on known benchmarks show that our solution is orders of magnitude more compact and significantly reduces network bandwidth. Several critical applications are enabled by TinyGarble, including but not limited to: (privacy-preserving) search, matching, genome computing, biometrics, as well as general purpose garbled processor in hardware for provably leakage-resilient cryptography.
You serve as an associate partner of the Intel Collaborative Research Institute for Secure Computing to aid in the development of solutions for the next generation of embedded secure devices. You have developed a way to foil hardware pirates using tiny physical variations between circuit elements on a chip--variations produced normally in the chip-manufacturing process. Can you tell our readers about your invention?
Microelectronic-based circuits and systems, which constitute the core technology for contemporary computing/communication in personal, industrial, business, and governmental affairs, are subject to adversarial acts. The standing problems include hardware piracy, malware, counterfeit products, and vulnerabilities of digital key storage (the root of most software and data security) in hardware. Before our work, tracing the pirated chips was not possible, since the ICs coming from the same mask all have identical digital functionalities.
Our research has invented the concept of active hardware metering, the first suite of security mechanisms and protocols that enable post-fabrication tracking and control of ICs during their life cycle. As a by-product of metering, we introduced the concept of logic encryption in hardware design, which opened vistas for research on this important topic. Our further research demonstrated that the technique can be adapted for authenticating the chip’s runtime states (and not just activation), interval licensing, and third-party IP authentication.
In your view, what is the biggest challenge with which your field is currently grappling?
In my view, the two biggest technical challenges in my field are: (1) the massive increase in the size of content and the dire need of intelligently processing data for information extraction, especially in real time, and (2) the extension of the boundary of the Internet to include a wide spectrum of non-conventional computing devices, known as the Internet-of-Things (IoT), as well as the emergence of intelligent vehicles and grids. In the new computing ecosystem of non-conventional devices and platforms, the attack domain is being further expanded to the physical world. In this setting, the vulnerabilities can cause catastrophic failures, and threaten critical infrastructures and people’s lives at massive scales.
Another major hurdle that, in my mind, is even more challenging to address is the training of a workforce that can fulfill the growing need for computing professionals. Training and bringing women to the computing fields is even more of a challenge. Meanwhile, this sector is expected to have one of the largest economic growths. From my perspective, these gaps are deeply concerning and if continued, will worsen the gender-based income inequality.
What is the biggest challenge that you have faced in your career?
The biggest challenge that I have faced in my career concerns gender biases that are still in place, even in an advanced country like United States today. Often times, I have to put exponentially more effort into proving myself worthy of a default status that a male colleague in similar standing would automatically obtain. What is heart-breaking is that some individuals are so biased that they don’t even see the fault in their approach.
Luckily, I have an excellent support system and confidence which helps me get by during such difficult times. I think my family’s belief in my capabilities and their attempt to foster my interests have instilled this confidence in me since childhood. My husband is also a constant source of inspiration and courage for me. I also have a great network of open-minded peers, colleagues, and mentors (both men and women); I rely on their advice in such moments of frustration. What worries me is that such biases can deter many other women, especially the younger and more sensitive ones, from the field. As a mother of a young girl and mentor to several others this is a problem that keeps me up at night. Through the new research center I’ve co-founded at UC San Diego – the Center for Machine-Integrated Computing and Security (MICS) – and by joining forces with similarly themed movements and networks, I hope to contribute to solutions addressing the complex, unfortunate root issues.
What research avenues are you exploring for the next few years?
The new UC San Diego research center that I have recently co-founded, MICS, is dedicated to addressing the two technical challenges in the field that I described earlier (in response to Question 6). Namely, we are focusing on bringing the automated machine-integrated solutions to address scalable and real-time data analytics and security problems. My collaborators and I are leveraging some of our recent results in this topic, but are simultaneously working on changing the paradigm and building novel solutions.
How did you initially decide to study Electrical engineering? How can prospective engineering students assess their skill and aptitude? What factors should prospective students consider when choosing an engineering school?
For me, the choice was straightforward. In Iran and in particular in Sharif University where I studied for my baccalaureate degree, it’s common for the top students to select electrical engineering or computer science as a major. And it turns out that I truly love the excitement and possibilities that are enabled by state-of-the-art technologies and the opportunity to create new concepts and methodologies. With the ever-increasing importance and permeation of computing across the board, the field will likely continue to enjoy tremendous growth for many years to come.
In terms of choosing a particular engineering school, it’s hard to give a generic advice since very person has a unique set of circumstances that may set boundaries on their choices. Generally speaking, there are certain perks to attending better-known programs. But as an undergraduate, if you make the right moves you have the ability to launch a successful life-long engineering career from any established undergraduate program. What’s important is learning the fundamentals, and also learning how to apply those fundamentals in real-world situations. No matter what school you attend, there will be many opportunities to join design teams, engineering competitions and other activities that allow you to apply, right away, what you are learning in your classes. Also, if you get involved in these activities as soon as possible, you are better positioned to get summer internships and research opportunities in faculty labs, which in turn make you a more competitive candidate for graduate programs and jobs after graduation.