Smart Metering and Hardware Trojan Electronic Digital Twin based on Artificial Intelligence

Modern electric power systems are based on the concept of smart grids. A key element of smart grids is smart metering, which is paramount for energetic load balancing and smart grid connections, particularly in industrial applications, being useful for the energetic diagnosis of production plants. In case of attacks on power meters, energy management procedures may be ineffective, with arising threats and damage to plants. This paper investigates the possibility of detecting wrong voltages in the industrial sensing field due to Hardware Trojan (HT) attacks. HT attacks involve a partial circuit modification generating wrong electrical signals and providing a wrong reading of the electrical power. This work simulates HT attacks through an open-source simulator, namely LTSpice, executing a parametric analysis of the circuital sensitivity by varying the HT resistance and capacitance that model the attack. Furthermore, the study proposes an Artificial Intelligence (AI) solution to reconstruct the voltage peak output after an HT attack. Hence, the combined implementation of the circuit layout and of the AI supervised algorithm performs an innovative Digital Twin (DT) suitable for HT industrial testbeds.