Optimizing Smart Home Intrusion Detection With Harmony-Enhanced Extra Trees

Abstract

In this study, we present an innovative network intrusion detection system (IDS) tailored for Internet of Things (IoT)-based smart home environments, offering a novel deployment scheme that addresses the full spectrum of network security challenges. Distinct from existing approaches, our comprehensive strategy not only proposes a model but also incorporates IoT devices as potential vectors in the cyber threat landscape, a consideration often neglected in previous research. Utilizing the harmony search algorithm (HSA), we refined the extra trees classifier (ETC) by optimizing an extensive array of hyperparameters, achieving a level of sophistication and performance enhancement that surpasses typical methodologies. Our model was rigorously evaluated using a robust real-time dataset, uniquely gathered from 105 IoT devices, reflecting a more authentic and complex network scenario compared to the simulated or limited datasets prevalent in the literature. Our commitment to collaborative progress in cybersecurity is demonstrated through the public release of our source code. The system underwent exhaustive testing in 2-class, 8-class, and 34-class configurations, showcasing superior accuracy (99.87%, 99.51%, 99.49%), precision (97.41%, 96.02%, 96.07%), recall (98.45%, 87.14%, 87.1%), and f1-scores (97.92%, 90.65%, 90.61%) that firmly establish its efficacy. This work marks a significant advancement in smart home security, providing a scalable and effective network IDS solution that is adaptable to the intricate dynamics of modern IoT networks. Our findings pave the way for future endeavors in the realm of cyber defense, ensuring that smart homes remain safe havens in an era of digital vulnerability.

Keywords

• Extra trees classifier
• harmony search algorithm
• hyperparameter optimization
• Internet of Things
• intrusion detection system
• machine learning