Cyber Attacks in the Internet of Things (IoT) and Intelligent Defense Mechanisms

Abstract

The Internet of Things (IoT) has connected billions of physical objects to the internet, enabling smarter homes, manufacturing, critical infrastructure, transportation, and healthcare. However, IoT ecosystems are increasingly vulnerable to cyberattacks due to the growing number of resource-limited devices, weak authentication techniques, and insufficient security deployments. Common threats include ransomware targeting smart firmware and home systems, Man-in-the-Middle (MITM) attacks intercepting device-to-device communication, and Distributed Denial of Service (DDoS) attacks exploiting compromised IoT devices as botnets. Because of their decentralized nature and limited computational resources, traditional security solutions such as firewalls and signature-based intrusion detection systems (IDSs) are not sufficient for IoT environments. Although many studies have addressed IoT security, a key gap remains in systematically reviewing and comparing recent intelligent and adaptive defense mechanisms for highly distributed and resource-constrained IoT systems. This study fills this gap by reviewing major IoT cyberattacks and analyzing intelligent defense strategies based on machine learning (ML), deep learning (DL), blockchain, and federated learning. ML and DL models are widely used for anomaly detection, traffic classification, and automated threat response, while blockchain enables secure authentication and decentralized trust management. Federated learning and lightweight encryption further support secure computation in constrained devices without significant performance loss. However, challenges remain in ensuring user privacy, achieving real-time detection, and balancing energy consumption with security requirements. The findings of this review show that proactive, cooperative, and self-healing security mechanisms are more effective than traditional reactive approaches. The study also provides recommendations for developing scalable, energy-efficient, and privacy-preserving IoT security frameworks for future research.