7.4 Using Classic Methods of Reliable …
689
testing procedures are performed on the IC model before manufacturing in silicon,
followed by testing after manufacturing in silicon. At the same time, the authors
proposed a mechanism for revising implementation and research by means of a
penetration test. Unfortunately, it remains unclear to us from the text of [72] how this
testing works to overcome protection without an exhaustive combination of message
flows. The HTDS method avoids this by adapting the ideology of “reinforcing learning” directly to the detection cycle, similarly to how is used in the classical theory
of machine learning.
Finally, here is a brief conclusion for those who simply looked through the text
of this section.
Hardware Trojans are a type of malicious circuits, which are unfortunately implemented in silicon today and turn out to be nothing less than harsh reality of everyday
life. While there are complex, practically verified, thought-through, and reliable
developments of software, there are little existing examples for safe implementation
of microcircuits.
We have already considered the original method of detecting hardware Trojans
suggested by the authors of [72], which can be effectively applied in the standard
microcircuit design route as well. Within the iterative cycle, rules and patterns leading
to a design rule set are extracted from a knowledge base. Automatic tests help identify
suspicious structures, the description of which is sent to the senior testing engineer
and security engineer (or their managers). These specialists are capable of tracing
these structures back and determining whether they are malicious or not. If the
structures are identified as malicious, subsequent critical analysis of the error is
performed, and the obtained information is transferred back into the knowledge base.
This detection cycle (HTDS) accompanies the microcircuit design route phases,
simultaneously expanding its knowledge base and adapting. Thus, it is taken into
account that certain properties of a malicious circuit are easier to find in created
objects during specific stages of development and methods of such detection and
are contained in the described method. This method is effective against design and
synthesis attackers and partly against fabrication attackers.
7.5 Using Sandbox as a Method of Protection
from Hardware Trojans in SoC
7.5.1 Introduction to the Problem
In the collective work [90] of the research laboratories of the US Air Force (cybersecurity department) and the University of Arkansas, one more effective approach
is presented to solve problems of protection against the introduction of hardware
Trojans into systems on chips (SoC). This approach is based on the fact that hardware Trojans pose a threat only when they are activated. Therefore, the essence of
solving the problem is to avoid labor-intensive and mostly destructive studies of
689
testing procedures are performed on the IC model before manufacturing in silicon,
followed by testing after manufacturing in silicon. At the same time, the authors
proposed a mechanism for revising implementation and research by means of a
penetration test. Unfortunately, it remains unclear to us from the text of [72] how this
testing works to overcome protection without an exhaustive combination of message
flows. The HTDS method avoids this by adapting the ideology of “reinforcing learning” directly to the detection cycle, similarly to how is used in the classical theory
of machine learning.
Finally, here is a brief conclusion for those who simply looked through the text
of this section.
Hardware Trojans are a type of malicious circuits, which are unfortunately implemented in silicon today and turn out to be nothing less than harsh reality of everyday
life. While there are complex, practically verified, thought-through, and reliable
developments of software, there are little existing examples for safe implementation
of microcircuits.
We have already considered the original method of detecting hardware Trojans
suggested by the authors of [72], which can be effectively applied in the standard
microcircuit design route as well. Within the iterative cycle, rules and patterns leading
to a design rule set are extracted from a knowledge base. Automatic tests help identify
suspicious structures, the description of which is sent to the senior testing engineer
and security engineer (or their managers). These specialists are capable of tracing
these structures back and determining whether they are malicious or not. If the
structures are identified as malicious, subsequent critical analysis of the error is
performed, and the obtained information is transferred back into the knowledge base.
This detection cycle (HTDS) accompanies the microcircuit design route phases,
simultaneously expanding its knowledge base and adapting. Thus, it is taken into
account that certain properties of a malicious circuit are easier to find in created
objects during specific stages of development and methods of such detection and
are contained in the described method. This method is effective against design and
synthesis attackers and partly against fabrication attackers.
7.5 Using Sandbox as a Method of Protection
from Hardware Trojans in SoC
7.5.1 Introduction to the Problem
In the collective work [90] of the research laboratories of the US Air Force (cybersecurity department) and the University of Arkansas, one more effective approach
is presented to solve problems of protection against the introduction of hardware
Trojans into systems on chips (SoC). This approach is based on the fact that hardware Trojans pose a threat only when they are activated. Therefore, the essence of
solving the problem is to avoid labor-intensive and mostly destructive studies of