Design And Detection Of Hardware Trojan Horse

The design and fabricate process of integrated circuit may be attacked by a hardware Trojan horse, which will bring potential security problems to the system using these ICs. This paper focuses on the design and detection of hardware Trojan horse, which has a significant meaning for integrated circuit. Firstly, the history, classification and detection of hardware Trojan are concluded. Secondly, the design methods of hardware Trojan are studied in detail, which will be very helpful to design a more effective hardware Trojan detection method. Finally, a new hardware Trojan detection strategy is proposed.The main contents of this paper include:1. For the low controllability disadvantage of the trigger circuit using a counter, this paper proposes a new trigger circuit based on up/down counter. The proposed circuit uses an external pulse as the trigger signal. If the trigger process fails due to some unexpected factors, the up/down counter will count in the reverse direction until it reaches the initial state. Then the trigger circuit will wait for the next trigger. This will improve the controllability and the anti-interference performance of the trigger circuit.2.Using an AES circuit as target circuit, a hardware Trojan inserted on the gate level based on fault attack is designed. This Trojan circuit occupies only four XOR gates. When it is triggered, an error occurs and the AES circuit outputs an error ciphertext. Using two pairs of correct/error ciphertext, the tenth round key can be cracked. Circuit simulation is performed and the round key is cracked by a matlab program.3. A hardware Trojan detection strategy based on ring oscillator is proposed. The whole detection process includes changing design, collecting data and detecting Trojan. The detection process and the configuration of ring oscillator have been described in detail. A python program is designed to analyze and change the circuit netlist. This paper uses the ISCAS-85:C432 circuit as the target circuit and performs simulation of the whole detection process on it. The simulation results prove the effectiveness of this detection strategy.