Information Sciences and Technology

Wu, Liu receive $499K to support system security research

UNIVERSITY PARK, Pa. -- System security is a foremost concern for many computer users, but current monitoring systems can interfere with the programs they are designed to protect. Dinghao Wu, assistant professor of information sciences and technology (IST), and Peng Liu, professor of IST, are developing an innovative security monitoring technology called “software cruisin4g3wsg” that detects software bugs and vulnerabilities without compromising a program’s effectiveness. To enable this research project, dubbed “Software Cruising for System Security,” the pair has received an award of $499,745 from the National Science Foundation (NSF).

“The goal of this project is to monitor security for programs by leveraging development in the computer hardware industry,” Wu said.

According to Wu, there has been a shift in the computer industry towards multi-core processor architectures. A multicore processor is a single computing component with two or more independent actual CPUs (called "cores"), which are the units that read and execute program instructions. While processors were originally developed with only one core, multicore processors are now widely used across many application domains including general-purpose, embedded, network, digital signal processing (DSP) and graphics. Multiple cores can run multiple instructions at the same time, increasing overall speed for programs amenable to parallel computing.

Runtime monitoring, a technique to enforce safety and security properties at program execution time, is essential to detect intrusions and keep the system healthy. However, Wu said, security monitoring enforcement often delays and blocks the execution of protected programs. Conventional concurrent runtime monitors have not been able to leverage the multicore architectures for performance due to synchronization issues. If conventional synchronization primitives are used, when the monitor is crashed or blocked due to external events, the protected program will also be blocked even if the monitor is not monitoring. The goal of Wu’s and Liu’s research is to explore multicore architectures for non-blocking concurrent security monitoring using lock-free data structures and algorithms.

The software cruising system, Wu said, would leverage multi-core architectures by “de-coupling” the computer program and the monitoring system so that they can run concurrently on a multicore computer. As a result, there would be much less risk that the monitoring system would interrupt any of the program’s functions. For example, if a hacker gains access to a computer, the system would raise an alarm to the user without altering the software.

“The key is to make this communication non-blocking,” Wu said.

Wu and Liu are in the beginning stages of a three-year project, Wu said. Eventually, they hope to deploy the technology to cloud computing systems and data centers that utilize thousands of computers. They are also planning for open source release in the near future.

 

Last Updated November 28, 2012

Contact