The Marconi Society community encompasses a wide range of technologists and advocates, from the prestigious Marconi Fellows to the advisors who are shaping our upcoming symposium in October. This network of leaders guides our work to improve global digital inclusion and diversity in STEM. 

Jennifer Rexford has served on the Marconi Prize Selection Advisory Committee since 2020, helping to evaluate and recommend our latest Marconi Fellow, Andrea Goldsmith. Rexford currently serves as the Gordon Y. S. Wu Professor in Engineering and Chair of Computer Science at Princeton University, where she has taught since 2005. Prior to her career in academia, Rexford worked for eight years at AT&T Research.

As a researcher and teacher, Rexford prioritizes a thoughtful, contextual approach to solving problems. She describes her areas of research interest as “Internet routing, network measurement, and network management, with the larger goal of making data networks easier to design, understand, and manage.” 

We sat down with Rexford to discuss her perspective on the “plumbing” of the Internet, the importance of building community, and the greatest challenges facing communications technologists.

How would you describe your approach to research and teaching?

My research focuses on how we can make the ‘net work from the point of view of the people who run it. I think of myself as an Internet plumber. Ultimately the research I do will benefit the end user, but my focus is on the network administrators and making the system more flexible, performant, and secure. When we talk about ICT innovations, we’re often focused on the applications and systems that make use of the underlying infrastructure of the Internet. I would like to see a future where the infrastructure itself is programmable and customizable. 

The beauty of the Internet is that it is connected to millions of programmable devices that allow for an amazing variety of innovation. But the inside of the network has not been very programmable. If we were able to make the fundamental structures programmable, we would be able to provide better transparency and more access to effect change on how the ‘net operates, creating “smart plumbing” that would be more agile in responding to security issues.

In my teaching, I think a lot about design patterns. It’s easy to get lost in the details and minutiae of networking issues, so I teach my students to see the patterns of problems and their solutions, which allows them to have a conceptual understanding of networking that can help solve problems across the board. The people who run networks have to master the details, but I want to train researchers who can approach network problems with a wider understanding of the “plumbing” of the Internet and how specific solutions might be applied in other cases.

How has the student population at Princeton changed over time? 

I’ve taught computer science at Princeton since 2005, and I’ve been lucky to teach a subject that is incredibly popular and continues to grow. Over the last decade, we’ve seen the population of computer science students diversify significantly. There are more women and underrepresented students in my courses than ever before. 

Students also bring a diversity of interest—10 years ago most of my students loved tech for its own sake, whereas many of my students now are interested in tech because they see its potential impact on social and humanitarian concerns. What they want to do with technology is all over the map. It’s challenging to teach them well since every student is a unicorn, but it’s incredibly rewarding to see their passion for an interdisciplinary approach to the field.

When I graduated, I didn’t have a sense of how my work and research would fit into the world. My students now have a strong sense of purpose and a confidence that if they understand computer science skills, they can do something useful and powerful when they graduate.

Your 1993 book, She’s an Engineer? Princeton Alumnae Reflect, collected comments of women alumni about Princeton’s engineering program. How has your perspective on being a woman in engineering changed since 1993?

At Princeton, about 40% of our computer science majors are women. It’s the second-most-popular major for women at the school. Once you have that critical mass of women students, you can create student groups that foster community and mentorship. We lose so many talented women in the early stages of the STEM pipeline (as early as middle school) because they don’t feel represented or encouraged. Building a community of women engineers—even high school students and undergraduate students—who can mentor those younger prospective computer science students is incredibly valuable.

Beyond gender diversity, we’ve worked really hard to make sure that the program is inviting and valuable for students from a variety of backgrounds. Our introductory computer science courses are designed to be exciting and accessible for students who have never taken a coding course before. 

You’ve advocated for the importance of a supportive, engaged community for any successful professional, and particularly those that are underrepresented in the field. How do you instill a sense of community in your students?

It’s been difficult to maintain that sense of community we have on campus through the pandemic. But the benefit to working remotely is that everyone’s personal life is much more visible, and that is very humanizing. My students see me balancing my family and my work every day. I think it’s good for students to experience those layers and understand that it’s possible to be a researcher, a teacher, a mentor, and a parent.

Even beyond COVID-19, academia can be a difficult context in which to develop and maintain community. Faculty members work with many students but peer collaboration is less common. I really value my relationships with my colleagues and most of my students are co-advised, which works well for Internet research because it touches so many other fields. It’s a natural fit for interdisciplinary work.

Why do you participate as a member of the Marconi Prize Selection Advisory Committee?

From a selfish point of view, it’s wonderful to be a fly on the wall and hear the members of the committee discuss the pressing issues of the field and the candidates we are considering. I love getting to know each candidate’s background. If I’ve only read their research, I might not be aware of the breadth and well-roundedness of how their work impacts humanitarian and entrepreneurial concerns.

Serving as an advisor for the Marconi Prize is rewarding because the award inspires the next generation of researchers. Each awardee is someone who has used their technical skills to have a great impact on the world. As a culture we often focus on technical excellence or humanitarian impact—this award shows us that it’s important to strike a balance between both. Awards help define what it means to be a leader in the field.

What is the greatest challenge in Information and Communications Technology (ICT) in the upcoming decade?

As a researcher, I am interested in improving the infrastructure of the network. A major challenge of my work is that we can’t simply turn off the Internet while we implement and test our changes. And while we think of the Internet as one system, it’s a network of networks, all of which are under different party’s control. It crosses economic, social, and national boundaries, and some of those systems have distinct and sometimes competing requirements for how it operates. It’s a very complex system.

Because there are so many intersecting systems and protocols that make the Internet run, it is also very technically difficult to find the right solution that will work universally. This is why I’m interested in making the Internet plumbing more programmable—a solution for my local community’s problem may not work in yours. Right now, we all live with this base system that is very difficult to change. 

One of the great democratizing aspects of the Internet is that if you have the skills to program a computer you can get your information out into the world. If we make the infrastructure more programmable, we can fundamentally expand the set of people who can impact it. The more we make the infrastructure flexible, the more we can create custom solutions informed by local concerns. It’s a very complicated project, but it would have a major impact on access, transparency, and flexibility of the network.