The detection method of probable areas of hardware Trojans location in FPGA-based components of safety-critical systems

摘要

The life cycle model of FPGA-based systems founded on emphasizing the stages of formation and modification of their program code is considered. At the stages next to the ones of low-level program code (bitstream) formation its integrity can be monitored with the help of hash embedded into the program code in the form of a digital watermark. Such kind of monitoring increases the complication of malicious logics fragments (hardware Trojans) implantation into the program code. At the earlier stages another (more formalized) set of actions is required for the counterstand to hardware Trojans emergence. In the given article the detection method of probable areas of hardware Trojans location in FPGA safety-critical systems is offered. The method is based on the fact that input data have different natures under the normal and emergency modes in safety-critical systems. The method offered can be used as the initial starting basis for counteraction to hardware Trojans implantation in the case when the program code integrity monitoring with the help of hash is impossible.