688
7 Countermeasures Against Hardware Trojans
sets of rules for detection of hardware Trojans. Therefore, it was designed as a feedback loop: in the process of its use, it accumulates new pieces of knowledge, updates
its rules and becomes more and more perfect over time. Moreover, many studies
and verification procedures can easily be automated. Therefore, it rightly complements previously known methods, such as [74], and partially uses the approaches
of [79]. Sets of rules and knowledge base shall be formed and used in cooperation
with industrial semiconductor production companies, similarly to the way international authoritative companies engaged in counteracting PC viruses (examined in
Chaps. 3–4) jointly used their accumulated knowledge and characteristic features of
various malicious inclusions.
7.4.8 A Brief Overview of Studies Similar to HTDS
Over the last 10 years, several papers were presented in scientific periodical press
that discussed methods of detection of circuitry Trojans similar to HTDS in terms of
ideas, e.g., the use of formal verification, i.e., checking equivalence of two different
representations of a design, leads to the method of structural testing [80, 81]. Other
similar close methods were aimed at activating the Trojan embedded in the IC,
detecting rare events that serve as a trigger for them. A number of other additional
methods make it possible to reduce the total amount of testing works, e.g., by applying
statistical analysis or meshing of the state space [82, 83]. It is also possible to compare
the measured physical parameters of the analyzed chip to the parameters of the
reference chip that definitely contains no Trojans (reference model), which helps to
detect side channels [84–86]. Another close method known as invasion consists in
inserting additional elements in the microcircuit design without altering its original
functionality in order to test it after serial production [84, 87]. The point of interest
here is the method for detecting their location in combination with the task of Trojan
localization (detection). In this regard, we can use the word “activation”, which means
the most intensive operation of the Trojan with simultaneously reduced activity of
the remaining parts [88-89]. Another group of methods indicated as mensuration
refers to obtaining non-standard information locally, from measurement of sidechannel parameters [85, 86]. All these works are aimed at various technical aspects
of detecting Trojans and strategies of mitigating the consequences of their attacks.
However, as of now, all this knowledge, to deep regret of the authors of this book, has
not been integrated in real development processes, which is objectively necessary to
counter Trojans in an organized manner. Clearly, the lifecycle of the HTDS method
examined here makes it possible to integrate them in a truly systemic way.
The work [74] by Khattri et al. also presents the lifecycle of a reliable method
of microcircuit design. Indicating the lack of adequate means for statistical analysis
of hardware description languages and the exhaustive set of hardware threats, the
authors of [74] also designed such a method consisting of five phases based on the
practical experience from reliable programming technique. Initial reliability assessment is accompanied by revision of architecture and revision of design. After that,
7 Countermeasures Against Hardware Trojans
sets of rules for detection of hardware Trojans. Therefore, it was designed as a feedback loop: in the process of its use, it accumulates new pieces of knowledge, updates
its rules and becomes more and more perfect over time. Moreover, many studies
and verification procedures can easily be automated. Therefore, it rightly complements previously known methods, such as [74], and partially uses the approaches
of [79]. Sets of rules and knowledge base shall be formed and used in cooperation
with industrial semiconductor production companies, similarly to the way international authoritative companies engaged in counteracting PC viruses (examined in
Chaps. 3–4) jointly used their accumulated knowledge and characteristic features of
various malicious inclusions.
7.4.8 A Brief Overview of Studies Similar to HTDS
Over the last 10 years, several papers were presented in scientific periodical press
that discussed methods of detection of circuitry Trojans similar to HTDS in terms of
ideas, e.g., the use of formal verification, i.e., checking equivalence of two different
representations of a design, leads to the method of structural testing [80, 81]. Other
similar close methods were aimed at activating the Trojan embedded in the IC,
detecting rare events that serve as a trigger for them. A number of other additional
methods make it possible to reduce the total amount of testing works, e.g., by applying
statistical analysis or meshing of the state space [82, 83]. It is also possible to compare
the measured physical parameters of the analyzed chip to the parameters of the
reference chip that definitely contains no Trojans (reference model), which helps to
detect side channels [84–86]. Another close method known as invasion consists in
inserting additional elements in the microcircuit design without altering its original
functionality in order to test it after serial production [84, 87]. The point of interest
here is the method for detecting their location in combination with the task of Trojan
localization (detection). In this regard, we can use the word “activation”, which means
the most intensive operation of the Trojan with simultaneously reduced activity of
the remaining parts [88-89]. Another group of methods indicated as mensuration
refers to obtaining non-standard information locally, from measurement of sidechannel parameters [85, 86]. All these works are aimed at various technical aspects
of detecting Trojans and strategies of mitigating the consequences of their attacks.
However, as of now, all this knowledge, to deep regret of the authors of this book, has
not been integrated in real development processes, which is objectively necessary to
counter Trojans in an organized manner. Clearly, the lifecycle of the HTDS method
examined here makes it possible to integrate them in a truly systemic way.
The work [74] by Khattri et al. also presents the lifecycle of a reliable method
of microcircuit design. Indicating the lack of adequate means for statistical analysis
of hardware description languages and the exhaustive set of hardware threats, the
authors of [74] also designed such a method consisting of five phases based on the
practical experience from reliable programming technique. Initial reliability assessment is accompanied by revision of architecture and revision of design. After that,