A measurement framework that uses micro-benchmarks to capture and catalog program-dependent signals at specific frequencies
Milos Prvulovic and Alenka Zajic from the School of Interactive Computing at Georgia Tech have developed a measurement framework that uses micro-benchmarks to capture and catalog program-dependent signals at specific frequencies. The end-goal of these inventions lies in preventing potential attackers from penetrating and obtaining sensitive/secret information within the computer. The first invention comprises of a methodology to rapidly discover activity-modulated signals using specially designed micro-benchmarks.
It utilizes recognizable spectral patterns that can identify those signals exhibiting a particular type of behavior such as amplitude modulation. The second invention, called CAMELIA, provides accurate and precise security monitoring of embedded, mission-specific, and traditional computing devices and software while keeping the monitoring and computing systems separate. CAMELIA uses the previous invention’s capability of measuring small signal emanations as the basis of security monitoring activity. This approach is revolutionary in that the monitoring is occurring outside the computing device or system versus current monitoring systems that reside within the computing device. Consequently, no modifications need to be made to the computing device to enable this security mechanism.
Accurately and precisely perform security monitoring of information and data being used by computer devices and systems
Eliminates the need for the security mechanism to be incorporated into the device/system being monitored
Capable of functioning without using resources of the device or system being monitored
Can be used as a means to benchmark and compare the performance of different computing devices used to handle sensitive data
These inventions can be applied in the areas of security monitoring, software profiling, code debugging and overall security enhancement of the device or system under consideration. Other application areas include embedded and mission-critical devices or systems where currently used security monitoring and attack prevention hardware and software tends to be more expensive, time consuming and adds to the overall size and form-factor of the device.