CAREER: Implementing a New Parallelization Model for Virtual Machines
Project Award Date: 03-01-2010
The multiyear Faculty Early Career Development (CAREER) award will support ongoing efforts to build more secure and better performing software systems.
ITTC researchers are developing a more secure and efficient framework for virtual machines (VMs), which ensure compatability between applications and the devices running them. Cell phones, PDAs and computers are among the billions of devices that have VM software running Internet programs and applications. To limit the cost and start-up time, current VMs apply only basic security checks. Devices are then left vulnerable to viruses and other malicious software that can corrupt and steal private data--from passwords to address books.
The new VM framework will slice out the security management and program monitoring tasks and perform them simultaneously with the main program. The framework will allow developers to provide new and more expensive security checks while minimizing the performance penalty incurred at runtime. The ITTC approach will employ program slicing to construct only the program state required for each security task. The proposed framework will naturally exploit the anticipated growth in the number of processing cores on a chip to run individual program slices concurrently with each other and with the main program thread. The framework will reduce the overhead of monitoring and security tasks and allow more secure and efficient execution of future programs.
Faculty Investigator(s): Prasad Kulkarni (PI)
Student Investigator(s): Hemaiyer Sankaranarayanan, Xing Liu, Marianne Jantz, , Michael Jantz, Donovan Carson, Jason Gevargizian, Elizabeth Alonzi, Surya Tej Nimmakayala, Forrest "Jamie" Robinson, Katherine Wu, April Wade, Monica Shafi, Dane Mitchell, Jacob Hegna
Primary Sponsor(s): National Science Foundation