چكيده به لاتين
Today, integrated circuits have different applications in different domains such as military industries, economic,etc . So the consequences of an attack can be expensive which indicates the importance of hardware safety studies.
In this thesis a new heuristic method based on Genetic Algorithm is proposed which can generate test sets for hardware Trojan detection. Innovation of this study is to provide a proper fitness function for genetic algorithm that proves better evaluation. The parameters of controllability, observability and transition probability of rare nodes have been used in this fitness function. Also, instead of direct evaluation test vectors, rare nodes are evaluated.
According to simulation of past works and proposed method can be followed to this point that the proposed method has better performance than previous works. The generation time for test sets dramatically has been reduced. Results of the simulation indicate 80% reduction in generation time for test sets (on average), compared to the same works. On the other hand reduced generation time for test vectors has been associated with an increase in trigger coverage percent. For each benchmark circuits (with one exception), the trigger coverage experiences 2% increment. of course, if inserted Trojans are very hard to trigger, trigger coverage increases to 23% .This is due to the selection of a new fitness function for genetic algorithm.
Keywords: hardware Trojan detection, Genetic Algorithm, Trigger coverage percent
