SSH-BA: Safeguarding Smart Homes with Intelligent IoT Behavior Analysis Using Classification Techniques

Smart home technologies automate regular operations like temperatures and lighting reacting to voice commands, making life easier. Smartphones allow homeowners to remotely operate these devices and monitor and manage their homes in real time. Smart home devices use IoT to make life smarter, more effective, and more customized. Smart home IoT devices may be hacked to access security cameras and personal data. Attackers might manipulate smart locks or thermostats, threatening people and their property. These IoT devices are networked, making strong cybersecurity necessary to guard against emerging threats. Early cybersecurity identification enables quick reaction and remediation, limiting damage and averting attack escalation. Early detection of security breaches allows proactive actions to stop unwanted access, limit the effect, and protect critical data. It dramatically minimizes the window of vulnerability, improving cybersecurity resilience and safeguarding systems from protracted exploitation or criminal activity. This research improves smart home security by training Aposemat IoT-23 dataset. XGBoost, Simple Neural Network, GaussianNB and Logistic Regression are evaluated to identify and prevent harmful data injection and illegal access. The results show they protect smart home networks with over 99% accuracy. Smart home IoT device protection is excellent with machine learning techniques like XGBoost and Simple Neural Network. These algorithms identify and block unwanted access and harmful data injection.