Dave Eckhardt's Research Interests

Areas of Interest

Operating systems
Computer networks & wireless networks
Distributed systems
File systems

Collaboration Available

While my primary focus at present is on teaching, I am interested in working with students on projects in the following areas:

Automatic configuration of link-level retransmission
Peer-to-peer file system architecture
Distributed operating system structure (mostly in the context of Plan 9).

Completed Projects

Membership on committees of completed thesis projects

Ph.D.

Ben Blum. Practical Concurrency Testing, or: How I Learned to Stop Worrying and Love the Exponential Explosion. CMU-CS-18-128 (2018).
Greg Hartman. Attentiveness: Reactivity at Scale. CMU-ISR-10-111 (2010).

M.S.

Qingyang Li. User Level Page Faults CMU-CS-20-124 (2020).
Sanjay Chandrasekaran. Towards a Low-Memory-Footprint, Container-Based IoT Security Gateway CMU-CS-19-117 (2019).
Da-Yoon Chung. OpenISR 2.0. CMU-CS-15-125 (2015).
Ben Blum. Landslide: Systematic dynamic race detection in kernel space. CMU-CS-12-118 (2012).
Adam Wolbach. Improving the deployability of Diamond. CMU-CS-08-158 (2008).
Ajay Surie. Improving mobile infrastructure for pervasive personal computing. CMU-CS-07-163 (2007).
David J. Matsumoto. A hybrid approach to Effort-Limited Fair scheduling in 802.11 (2003).

Ph.D. Dissertation

In May of 2002 I defended my Ph.D. dissertation, titled An Internet-style Approach to Managing Wireless Link Errors. Briefly, I investigated some of what must be done to extend the "Internet experience" across error-prone wireless links. In particular, I argued that protocol-blind link-level adaptive error control is the right solution for bursty packet errors. I then proposed a new notion of fairness for flows traversing noisy links, effort-limited fairness (ELF), which ensures sensible outcomes in response to link capacity loss. Evaluation was based on replaying bit-level error traces past a prototype implementation running in a NetBSD kernel.

Credit Net

Many moons ago, I was part of the Credit Net project.