As the process of economic globalization deepens,a large variety of integrated circuit designers tends to move the chip's manufacturing to the developing country.But during the globalization of semiconductor desig...As the process of economic globalization deepens,a large variety of integrated circuit designers tends to move the chip's manufacturing to the developing country.But during the globalization of semiconductor design and fabrication process,integrated circuits are suffering from increasing malicious alterations from the untrusted foundries,which pose a serious threat to the military,finance,transportation and other critical systems.An noninvasive approach was presented to measure the physical "sidechannel"parameter of a chip such as current or delay,which is effectively capable of detecting the malicious hardware alternations.The intrinsic relationship of a circuit's side-channel parameters was exploited to distinguish the effect of a Trojan in the presence of large random noise and process noise,such as the Dynamic current(IDDT)versus the maximum operating frequency(Fmax)correlation.The Monte Carlo analysis in Hspice using ± 20% Gauss variations in transistor threshold voltage(Vth)was carried out to simulate the circuit state in the real world.Simulation Results show that this approach is effective to detect the ultra-small Trojans.展开更多
文摘As the process of economic globalization deepens,a large variety of integrated circuit designers tends to move the chip's manufacturing to the developing country.But during the globalization of semiconductor design and fabrication process,integrated circuits are suffering from increasing malicious alterations from the untrusted foundries,which pose a serious threat to the military,finance,transportation and other critical systems.An noninvasive approach was presented to measure the physical "sidechannel"parameter of a chip such as current or delay,which is effectively capable of detecting the malicious hardware alternations.The intrinsic relationship of a circuit's side-channel parameters was exploited to distinguish the effect of a Trojan in the presence of large random noise and process noise,such as the Dynamic current(IDDT)versus the maximum operating frequency(Fmax)correlation.The Monte Carlo analysis in Hspice using ± 20% Gauss variations in transistor threshold voltage(Vth)was carried out to simulate the circuit state in the real world.Simulation Results show that this approach is effective to detect the ultra-small Trojans.
