CityShield: An IoT-Driven and AI-Based Threat Detection System for Smart City Operations
Md. Asif Sarker Emon, Md. Faruk Abdullah Al Sohan, Md. Maruf Hossain Munna, Mustakim Ahmed, Kazi Redwan · 2025
In this era of technology, smart cities are becoming more common, but their growing connectivity exposes them to continuous cybersecurity threats, putting their operations and security at risk. This research focuses on enhancing cybersecurity in smart cities by implementing a real-time threat detection and response system, with key metrics such as faster detection speeds, reduced false alarms, and improved system adaptability to evolving threats. Robust safety protocols are required because attacks by ransomware, breaches of information, and weak IoT devices endanger vital services. This research proposes a system that presents the implementation of real-time data collection through IoT devices and a streamlined data flow leveraging Kafka for rapid ingestion, HDFS for scalable long-term storage, and Apache Spark for efficient processing. The proposed system uses Autoencoder algorithms to detect anomalies and DQN to make fast, intelligent decisions responding to detected threats. The proposed system aims to strengthen clever city defenses by continuously monitoring various city services. Robust and adaptable threat detection and response systems that can re-duce vulnerabilities are guaranteed through this technological integration. This versatile system protects vital infrastructure by changing it to counter new cyber threats. It enhances city safety and strengthens smart city cybersecurity by detecting threats and safeguarding critical information and infrastructure for a secure future. With this proposed strategy, smart cities could become safer, more technologically advanced, and more resilient. This proposed solution allows smart cities to become technologically advanced, secure, and resilient.