*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Abstract:
Many techniques for optimization, debugging, and security rely on obtaining detailed information about the program's execution. Unfortunately, existing approaches for collecting this information rely on heavily modified software and/or hardware of the program and/or system we are trying to observe.
Worse, such monitoring suffers from “the Observer's Effect,” which enables observable changes in the behavior of the system we are trying to observe. This is especially problematic when the purpose of observation is to gather fine-grain timing information; systems whose behavior adapts to meet real-time demands.
In this talk, I will describe our work with electromagnetic (EM) emanations from monitored system to obtain fine-grin information about program execution, without any modification to the observed system or its software, thus removing Observer's Effect. As two examples of this approach, I will present ZOP (Zero Overhead Profiling), our technique for recovering the control flow path taken by the program, along with the actual timing of control flow events, and Spectral Profiling, our low-cost technique for profiling and verification of loop execution. Finally, I will discuss the new opportunities and research challenges in our overall approach, and how we plan to address them in Camelia, our DARPA-funded project on using EM emanations to monitor the security of Internet-of-Things (IoT) devices.
Bio:
Milos Prvulovic is a professor in the School of Computer Science at the Georgia Institute of Technology.
His research focuses on hardware and software support for program monitoring, debugging, and security. His research on understanding the relationship between program execution and the resulting "side-channel" signals has led to widespread interest from professional societies, the media, and research sponsors – most recently earning a $9.4 million award from the Defense Advanced Research Projects Agency (DARPA).
In general, the goal of his research is to make both hardware and software more reliable and secure. Dr. Prvulovic is a senior member of the Association for Computing Machinery (ACM) and the Institute of Electrical and Electronics Engineers (IEEE), served as the chair of the IEEE Technical Committee on Microprogramming and Microarchitecture in 2016, and he is a member of the Steering Committee for the ACM/IEEE MICRO conference.