David Naylor

I make network architectures private and secure (without giving up everything else).

I'm a sixth year Ph.D. student in the Computer Science Department at Carnegie Mellon University, and I'm interested in computer networking and security. My advisor is Peter Steenkiste.


I received degrees in computer science and mathematics from The University of Iowa, where I worked with the UI Computational Epidemiology Group.


GHC 7509
School of Computer Science
Carnegie Mellon University


Managing Privacy Tradeoffs

Balancing privacy and {accountability, functionality, performance}.

Share Count AnalysisHotNets '15

Share Count Analysis is our first step toward a general privacy measurement methodology—that is, a technique for measuring how private an arbitrary network architecture or protocol is.

Multi-Context TLS (mcTLS)SIGCOMM '15

Multi-Context TLS allows endpoints to explicitly add trusted middleboxes to encrypted sessions. It provides fine-grained access control, so middlebox access can be restricted to read-only or limited to only part of the data stream.


Accountable and Private Internet Protocol (APIP)SIGCOMM '14

The Accountable and Private Internet Protocol (APIP) is designed to balance accountability and privacy. Senders are able to hide their addresses from the network by delegating responsibility for their packets to a trusted third party, which vouches for its clients' traffic unless a flow has been reported as malicious.

State of the Web

Monitoring the Web as we move towards HTTP/2 and ubiquitous HTTPS.

EyeorgCoNEXT '16

Eyeorg is a tool for crowdsourcing Web quality of experience measurements.

The Cost of the "S" in HTTPSCoNEXT '14

A collection of measurements showing the deployment and costs of HTTPS. Costs include handshake latency and loss of middlebox services like caching and compression.

Is the Web HTTP/2 Yet?PAM '15

eXpressive Internet Architecture

An Evolvable, Expressive, and Secure Future Internet Architecture

Demo at GEC15

In October 2012, we demoed XIA running on the GENI network at the 15th GENI Engineering Conference.

Computational Epidemiology

Measuring, modeling, and simulating disease spread.

Sensing Social Networks

Through my work with CompEpi, I helped deploy a sensor network in the intensive care unit at the University of Iowa Hospitals and Clinics. We gathered anonymized data telling us when healthcare workers were in contact with one another, when they were in contact with patients, and when they washed their hands. This data helped us answer questions about hand hygiene policy compliance and explore how communicable diseases might spread through the unit.

Scrub Scrub Revolution

Our sensor network project helped us monitor when healthcare workers washed their hands, but equally important is how. The World Health Organization has published recommendations for proper handwashing technique; these guidelines include a set of recommended scrubbing motions. I implemented a training “game” in which users scrub along with an expert while wearing accelerometers on their wrists, allowing the sytem to provide feedback on their scrubbing technique.