Advanced threat detection for smart cities through IoT sensor and network data integration with IoT-securefusion

As urban areas evolve into “smart cities,” driven by IoT devices that oversee various aspects such as traffic and public safety, the imperative for stringent security measures has intensified significantly. Conventional security systems generally depend on either network data (e.g., communication traffic) or sensor data (from devices observing real-world conditions); however, this methodology frequently results in vulnerabilities, mainly when attacks focus on specific components of the system or exploit the convergence of physical and digital layers. This paper presents IoT-SecureFusion, a novel two-phase security framework for smart cities. Our approach incorporates threat detection models that evaluate both IoT sensor data and network traffic, subsequently utilizing a decision fusion technique to amalgamate insights from both data sources into a cohesive, actionable security response. We assess IoT-SecureFusion utilizing a real-world dataset that integrates network traffic with sensor data from essential smart city systems, including traffic management, public safety, and utilities. We offer explicit labeling guidelines for standard operations and possible attacks to guarantee transparency and reproducibility. Our findings show that combining both data sources improves threat detection in IoT-SecureFusion, enabling the recognition of subtle attacks that traditional systems might miss. It enhances detection accuracy by 15% compared to using only network data and by 20% when solely relying on sensor data. This unified approach improves the safety of urban environments and allows for quicker, more effective reactions to dangers, preventing possible harm before it occurs.