Research Article

DHRCA: A Design of Security Architecture Based on Dynamic Heterogeneous Redundant for System on Wafer

Table 3

Experimental parameters and meanings.

ParameterValueMeaning

0.33Probability that the triggered HT in the chiplet is a function tampering HT
0.22Probability that the triggered HT in the chiplet is an information leakage HT
0.45Probability that the triggered HT in the chiplet is a denial-of-service HT
Probability that the outputs of two chiplets remain consistent after being tampered with
The average attack time of HTs is 100, 10 hr, and 6 min, simulating weak, medium, and strong attack scenarios
The average time the HTs was active is 100 hr and 6 min, simulating the long and short duration of the HTs
The average time for nonarbitration triggered dynamic switching chiplets is 100 hr, simulating the dynamicity of DHRCA
The average time for the chiplet to perform a task is 100, 1 hr and 6 min, simulating long time, medium time, and short time task
KProbability that TPAD runtime monitoring technology detects HTs when HTs are activated

Note: The parameter settings are based on the HT benchmarks and their distribution provided by the Trust-Hub [36] website. We have analyzed the chip-level HT benchmarks from Trust-Hub (100 in total) and categorized them based on their impact. Out of these benchmarks, there are 33 Trojans that tamper function, 22 HTs that leak information, and 45 HTs that cause denial-of-service.