On the optimization of Software Obfuscation against Hardware Trojans in Microprocessors

The quest of low production cost and short time-tomarket, as well as the complexity of modern integrated circuits pushed towards a globalization of the supply chain of silicon devices. Such production paradigm raised a number of security threats among which Hardware Trojan Horses (HTHs), that became a serious issue not only for academy but also for industry in the very last years. Indeed, it has been demonstrated that HTHs can be inserted into microprocessors allowing the attacker to run malicious software, to acquire root privileges or to steal secret information. In this paper we present the use of software obfuscation to protect systems against HTHs that aim at stealing information from the microprocessor while it is executing a program. Moreover, we present a Genetic Algorithm-based approach to optimize such anti-HTH methodology by maximizing the obtained obfuscation while minimizing the introduced overhead. We proved the effectiveness and efficiency of the proposed methodology on the Ariane 64bit RISC-V microprocessor running a set of MiBench benchmarks and cryptographic programs.