Quantum Optimization for IoT Security Detection

The digital transformation and innovation of today’s environment requires a focus on the security of systems to preserve their confidentiality, integrity, and availability. Considering these needs, in this work an efficient and high-performance intrusion detection system based on Quantum Annealing is presented. The goal is to identify attacks in an IoT environment such as Smart Home. The goal is to reduce the processing time for identifying attacks and to respond promptly to implement measures appropriate to the threat. We use Qboost, an iterative training algorithm in which a subset of weak classifiers is selected by solving a hard optimization problem in each iteration. The models were tested on a IoT Fridge dataset, characterized by a large volume of traffic with a low number of features and highly imbalanced data distribution. The results show an optimization of computational time and an improvement in detection accuracy.