7.4 Using Classic Methods of Reliable …
677
the existing methods of mitigation of such attacks (such as [74]) introduced additional manual verification during various process stages, but didn’t develop specific
effective methods of detection of circuit Trojans.
Based on their many years of practical experience with so-called classic technique
of reliable programming, the authors of [72] demonstrated that a fitting solution to
such problems associated with introducing malware on silicon can be the adoption
of similar methods for microcircuit design process as well. Due to significant differences between these fields of knowledge (microcircuit design and classic reliable
programming), mutual adaptation of these methods is not a trivial task. This is due
to the fact that a microcircuit leaves us with fairly limited possibilities in terms of
improving its security after its installation in an equipment piece. Moreover, attacks
on microcircuits are usually targeted, i.e., associated with a specific victim or a
product, while a malicious hardware implant (hardware Trojan) is usually aimed at
a wide range of potential anonymous victims.
Moreover, as follows from the previous chapter, drastic changes in the microcircuit
design process, which is well known to and understood by the developer, usually
require significant financial and organizational efforts, which is not always possible
to implement in practice. Therefore, the authors of [72] tried to optimally modernize
(change) these existing methods, adapting for this purpose only some of them, which
had been used earlier for development of trusted (safe) software.
As a first step, they analyzed known structures of typical malicious software in
order to formulate technical requirements for the task of successful detection of
hardware Trojans in microcircuits. After that, they analyzed flaws in the processes
of qualified circuit design and assessed reliable programming techniques in terms of
their applicability to the circuit design process. Methods recognized as successful
were later improved and included in design routes of microcircuit hardware.
Knowing the main types of potential attacks and their points of application, they
selected methods that are possible to implement and formulated requirements for
their correct application in microcircuit design. After that, these technical solutions
were accumulated as a special method of hardware Trojans referred to as hardware
Trojan detection Cycle (HTDC) by the authors. In particular, they demonstrated that
certain properties of Trojans are easier to detect during specific development phases.
It could be said that the HTDC method contains sets of rules that self-adjust to a
specific phase (stage) of IC design or production.
The HTDC method is based on the following principles [72]:
– identification of the specific threat model to formulate requirements to safe circuit
design process;
– introduction of complete traceability for development of the circuitry;
– determination of the lifecycle of the hardware Trojan in the IC to be considered
during design.
Taking into account the above, we will further consider in detail the features of
development of circuits, types of attacks, differences between microcircuit development and software development with maximum use of text and graphic materials of
the original [72].
677
the existing methods of mitigation of such attacks (such as [74]) introduced additional manual verification during various process stages, but didn’t develop specific
effective methods of detection of circuit Trojans.
Based on their many years of practical experience with so-called classic technique
of reliable programming, the authors of [72] demonstrated that a fitting solution to
such problems associated with introducing malware on silicon can be the adoption
of similar methods for microcircuit design process as well. Due to significant differences between these fields of knowledge (microcircuit design and classic reliable
programming), mutual adaptation of these methods is not a trivial task. This is due
to the fact that a microcircuit leaves us with fairly limited possibilities in terms of
improving its security after its installation in an equipment piece. Moreover, attacks
on microcircuits are usually targeted, i.e., associated with a specific victim or a
product, while a malicious hardware implant (hardware Trojan) is usually aimed at
a wide range of potential anonymous victims.
Moreover, as follows from the previous chapter, drastic changes in the microcircuit
design process, which is well known to and understood by the developer, usually
require significant financial and organizational efforts, which is not always possible
to implement in practice. Therefore, the authors of [72] tried to optimally modernize
(change) these existing methods, adapting for this purpose only some of them, which
had been used earlier for development of trusted (safe) software.
As a first step, they analyzed known structures of typical malicious software in
order to formulate technical requirements for the task of successful detection of
hardware Trojans in microcircuits. After that, they analyzed flaws in the processes
of qualified circuit design and assessed reliable programming techniques in terms of
their applicability to the circuit design process. Methods recognized as successful
were later improved and included in design routes of microcircuit hardware.
Knowing the main types of potential attacks and their points of application, they
selected methods that are possible to implement and formulated requirements for
their correct application in microcircuit design. After that, these technical solutions
were accumulated as a special method of hardware Trojans referred to as hardware
Trojan detection Cycle (HTDC) by the authors. In particular, they demonstrated that
certain properties of Trojans are easier to detect during specific development phases.
It could be said that the HTDC method contains sets of rules that self-adjust to a
specific phase (stage) of IC design or production.
The HTDC method is based on the following principles [72]:
– identification of the specific threat model to formulate requirements to safe circuit
design process;
– introduction of complete traceability for development of the circuitry;
– determination of the lifecycle of the hardware Trojan in the IC to be considered
during design.
Taking into account the above, we will further consider in detail the features of
development of circuits, types of attacks, differences between microcircuit development and software development with maximum use of text and graphic materials of
the original [72].