IOT-DRIVEN SMART CITIES: ENHANCING ATTACK DETECTION VIA CLOUD-BASED ANALYTICS AND MULTIFACTOR AUTHENTICATION

Abstract

Smart city infrastructures rely on extensive networks of Internet of Things (IoT) devices, which introduce new security challenges as sophisticated cyber threats target these connected systems. This paper proposes a cloud-enabled cybersecurity framework that combines edge-based anomaly detection with multifactor authentication to protect IoT-driven smart cities. In the proposed approach, resource-constrained edge devices (Raspberry Pi nodes) perform local data collection and preliminary analysis, while the heavy computation of attack detection is offloaded to the Amazon Web Services (AWS) cloud for scalability. A machine learning model (Artificial Neural Network) analyzes IoT sensor and device usage patterns in real time to identify anomalies indicative of attacks such as Distributed Denial of Service (DDoS). Also, a robust user authentication mechanism is integrated, employing biometric verification (e.g., facial recognition) coupled with one-time password (OTP) validation via the “Twilio” API. This two-factor authentication ensures that only authorized users can access or control critical IoT resources. The system’s performance was evaluated through a prototype smart home/city environment. The results demonstrate improved attack detection accuracy and low false alarm rates, while the multifactor authentication achieved high reliability with minimal latency. By leveraging an AWS cloud backend and “Twilio”-based OTP delivery, the framework enhances the overall security posture of smart cities.