Hicks photo

Michael Hicks
E-mail:
Twitter: @michael_w_hicks
Phone/Fax: +1-301-405-2710 / +1-301-405-6707
Office:

5246 Iribe Center

(Building 432 in the linked map)
Address:

Dept. of Computer Science

8125 Paint Branch Dr

University of Maryland

College Park, MD 20742


Publications PL Enthusiast (Blog)
CMSC 330
 MOOC on Software Security

I am a professor in the Computer Science Department and UMIACS at the University of Maryland, College Park. With David Van Horn I direct PLUM, the lab for Programming Languages research at the University of Maryland. I am also affiliated with the Maryland Cybersecurity Center (MC2), and was formerly its Director (see our video!). You may find it interesting to read about how we manage PLUM.

Here is my current vita and a list of my publications, organized by year and by category.

I received my Ph.D. in Computer and Information Science from the University of Pennsylvania in August 2001, and I spent one year as a post-doctoral associate affiliated with the Information Assurance Institute of the Computer Science Department at Cornell University. During academic 2008 - 2009, I was on sabbatical in Cambridge, England. From September to November I was at Microsoft Research and from December to August 2009 I was at the University of Cambridge Computer Laboratory. I was the director of MC2 from October 2011 to 2013. During the Summer of 2015 I visited Microsoft Research in Redmond.

How fast can you type? (My best so far is 108 wpm.)

Research

My primary research interest is to develop and evaluate techniques to improve software availability, reliability, and security. I am currently working on a number of projects.
Secure programming - How do we build software that is secure? We have been developing a contest, called build-it, break-it, fix-it whose aim to test how well students can build software securely. We have now offered the contest several times (including many times in classes at UMD and elsewhere) and have some interesting data analysis of what strategies work and don't work for building and breaking. We are working on a more in-depth analysis of contest data, and plan controlled studies to further test hypotheses. One clear outcome, previously known but confirmed by the contest, is that programming in C and C++ is risky. I have been working with Microsoft on Checked C, an extension to C aimed to provide safety. Checked C is specifically designed to ease the porting of legacy code.
Protecting against information leakage - How do we avoid leaking information via observable behavior? We are developing novel static analyses and tools to identify when a program might be leaking sensitive information via its direct outputs, running time, space usage, address/IO trace, and more. We have recently developed LWeb, a novel Haskell-based framework for preventing leaks in web applications. We are also exploring the use of fuzz testing. Finally, we have implemented approaches using probabilistic abstract interpretation for enforcing quantitative knowledge-based policies that protect static and time-varying secrets, and have recently looked at uses of sampling to improve precision.
Quantum computation programming languages - means to develop reliable and efficient quantum programs on near-term devices. We are looking at the challenge of developing high-quality quantum programs despite limited, noisy resources on quantum computers available in the near term. We are applying ideas from program verification and approximate computing. 
Blending programming languages and cryptography - means of implementing privacy-preserving or integrity-assuring computation through the combination of programming languages and cryptographic techniques. I have looked at languages and analyses for secure multiparty computation, most notably a new programming language called Wysteria, and a follow-on language Wys* for proving properties of multi-party computations. I have also developed novel mechanisms for cloud-based computations involving general-purpose authenticated data structures and compiler-optimized oblivious RAM.
Dynamic software updating (DSU) - means to safely, efficiently, and flexibly update running code. We have developed general-purpose methods for updating C and Java programs, and are currently considering means to update controllers in software-defined networks. Our system for dynamically updating C code is called Kitsune and our system for Java is called Rubah; the code and benchmarks for both are freely available. I have recently considered DSU for key-value stores and software-defined networks. I have also worked to improve the reliability of dynamic updates through application of multi-version execution. Earlier work and papers are described here.

Previously, I was involved in the following projects:

  • Adapton, a library for incremental computation --- the idea is to write an algorithm largely as usual, but then to derive a version that can incrementally update the output following a small changes to its input.
  • Expositor, a library for writing dynamic analyses to assist in debugging, taking advantage of record/replay support.
  • Diamondback Ruby, static and hybrid static/dynamic type system for the Ruby scripting language. 
  • Otter, a symbolic executor for C programs.
  • LockSmith, a static analysis tool for proving the absence of race conditions in C programs.
  • Path Projection, a browser-based UI toolkit for presenting, navigating, and querying paths emitted as static analysis results; we applied it to Locksmith.
  • Cyclone, a safe dialect of C. Cyclone's system for manual memory management was influential in the development of Rust.
  • I have also looked at means for customized, language-enforced security policies, implemented in a web programming language, SELinks, and automatically inserted by a compiler called Coco.
  • MGRP, for measurement-aware data transport and kernel-based rootkit detection.

Links to all past projects may be found on the PLUM home page.

Research Group

Current students and postdocs:
Kesha Hietala
Ian Sweet
James Parker
Leonidas Lampropoulos (postdoc, co-advised with Benjamin Pierce)
Robert Rand (postdoc, co-advised with Xiaodi Wu)

Previous students and postdocs:
Andrew Ruef Tools and Experiments for Software Security
Researcher at IDA/CCS since February 2019
Chang Liu**** Trace Oblivious Program Execution
Post-doc at UC Berkeley, 2016-2018; researcher at Citadel Securities since 2019
Shiyi Wei (postdoc) Assistant Professor, University of Texas at Dallas, since August 2017
Aseem Rastogi Language-based Techniques for Practical and Trustworthy Secure Multi-Party Computations
Researcher at Microsoft Research India since June 2016
Matthew Hammer (postdoc) Research scientist at DFinity since January 2019; previously Assistant Professor, University of Colorado, Boulder, August 2015-January 2019
Luís Pina*** Practical Dynamic Software Updating (for Java)
Assistant Professor at University of Illinois, Chicago, starting Fall'19; previously post-doc at George Mason (Aug 2017-19) and Imperial College, London (Mar 2015-Aug 2017)
Karla Saur* Dynamic Upgrades for High Availability Systems
Senior Software Engineer at Microsoft since October 2018; previously researcher at Intel Labs September 2015-2018
Piotr Mardziel Modeling, Measuring, and Limiting Adversary Knowledge
Systems Scientist at CMU (previously, post-doc) since June 2016; post-doc at UMD Jan'15 - Jun'16
James Parker (MS) LMonad: Information Flow Control for Haskell Web Applications
Research programmer at UMD January 2015-2019, now PhD student at UMD, part-time software research engineer, Galois
Nate Parsons (MS)
 Implementing and Typing a Core Calculus for Mixed-mode Secure Multi-party Computations (scholarly paper)
Missions software engineer at Planet since 2013; previously, engineer at JHUAPL
Khoo Yit Phang* User-centered Program Analysis Tools
Principal Software Engineer at MathWorks since August 2013
Nataliya Guts (postdoc)
Chris Hayden* Clear, Correct, and Efficient Dynamic Software Updates
Software Engineer at SocialCode since Dec. 2015 (at WaPo Labs/Trove 2012-2015)
Ted Smith (undergrad)* Insight Deliveryperson, Bloomberg LP, since 2019; Software Engineer, Google, 2016-2018; grad student at UMass Amherst 2013-2016
Stephen Magill (postdoc) Principal scientist, Galois, since 2014 (at IDA/CCS, 2012-2014)
Justin McCann Automating Performance Diagnosis in Networked Systems
Avere Systems since July 2012
Martin Ma* Improving Program Testing and Understanding via Symbolic Execution
Software Engineer at Google since 2013 (previously at Amazon)
Saurabh Srivastava* Satisfiability-based Program Reasoning and Program Synthesis
Founder, Synthetic Minds, since 2017; founder, 20n, 2013-2017; post-doc at Berkeley 2012-2014
Pavlos Papageorgiou The Measurement Manager: Modular and Efficient End-to-end Measurement Services
Software Engineer, Google (AI), since December 2008
Iulian Neamtiu Practical Dynamic Software Updating
 Assoc. Prof, NJIT, since Fall 2015 (at UC Riverside, 2008-2015).
Manuel Oriol (postdoc) Principal Scientist, R&D manager at ABB Switzerland Ltd. since Fall 2011;
Senior Lecturer, University of York (UK), 2008-2014
Polyvios Pratikakis* Sound, precise, and efficient static race detection for multithreaded programs
Assistant professor in CS, University of Crete, since 2014; researcher, Institute of Computer Science, FORTH, 2010-2014; post-doc at CNRS/VERIMAG 2008-2009
Nikhil Swamy Language-based Enforcement of User-defined Security Policies as Applied to Multi-tier Web Applications
Senior Researcher, Microsoft Research, Redmond, since Fall 2008
Nick L.  Petroni** Property-based Integrity Monitoring of Operating System Kernels
Chief scientist, Volexity, since 2015; research scientist, IDA/CCS 2008-2015

* co-advised with Jeff Foster ** co-advised with Bill Arbaugh *** co-advised with Luís Veiga
 **** co-advised with Elaine Shi

I have also worked closely with Niki Vazou (postdoc supervised by David Van Horn), David Darais (advised by David Van Horn), Andrew Miller (co-advised with Jon Katz and Elaine Shi), Avik Chaudhuri, Mike Furr, David An, and Elnatan Reisner (advised by Jeff Foster), Adam Bender (advised by Bobby Bhattacharjee), Jaime Spacco (advised by Bill Pugh), and Suriya Subramanian (advised by Kathryn McKinley while at UT Austin). I have previously advised Willem Wyndham, Jonathan Turpie (now at Amazon), Brian Corcoran (now at Palantir), Eric Hardisty, and James Rose (now at Google). I've also worked with post-grad Patrick Jenkins, undergrad Jeff Meister, and high school students, Ted Smith (from Walt Whitman High), and George Kleses, Matt McCutchen, and Cody Burton (from Montgomery Blair). Both Ted and Matt later became undergraduate students in our Department and Ted, Matt, and Cody all went to graduate school (at UMass, MIT CSAIL, and MIT Physics, respectively).

Teaching

Professional Activities

I am the past Chair of ACM SIGPLAN (served July 2015-2018); I was an Associate Editor for TOPLAS from 2012-2016; and I have served (or am serving) on the following committees

2019 POPL SRC, ASPLOS (ERC), SNAPL, CCS (AC), SecDev
2018 POST, S&P (AC), SecDev
2017 S&P, ESSoS, CCS, USENIX ASE, SecDev
2016 PLDI, CSF (PC co-chair), USENIX ASE, SecDev (PC chair), CSAW judge
2015 S&P, CSF (PC co-chair), SNAPL
2014 OBT, CSF, OOPSLA/SPLASH
2013 POPL (ERC), PLDI
2012 POPL (program chair), HotSWUp
2011 TLDI, HotSWUp (co-organizer), OOPSLA
2010 ESOP, PLDI (ERC and tutorials chair), ICFP (PC and local arrangements), PASTE
2009 POPL, S&P, PLDI SRC
2008 CCS, CATARS, COORDINATION, ISMM (ERC)
2007 PLAS (general and program chair), OOPSLA, COORDINATION, PLDI
2006 FTfJP, PLAS, SPACE, OOPS (part of SAC 2006)
2005 SCOOL, VEE
2004 IWAN, ICPP, FUSE
2003 IWAN, USE
2002 IWAN, USE
2001 IWAN
ATYlogo Support Wikipedia
Web Accessibility