Internet of Things Security: Emerging Threats, Authentication Systems, and Privacy-Preserving Mechanisms

Abstract

The geometric expansion of the Internet of Things (IoT), projected to exceed 60 billion connected devices by 2030, has fundamentally transformed industrial, civic, and medical infrastructures. However, this proliferation has introduced a fragmented attack surface characterized by an average of 5,000 attacks per month per device and projected global cybercrime costs of $10.5 trillion by 2025. This paper examines the shift from traditional security to sophisticated, AI-driven threat landscapes, including autonomous threat agents and large-scale botnet campaigns like BadBox 2.0. We analyze emerging technical defenses, specifically Lightweight Blockchain Authentication Models (LBAM) that reduce computational costs by 70%, and Privacy-Preserving Mechanisms such as Federated Learning (FL) combined with Homomorphic Encryption. Furthermore, the study explores the integration of Zero Trust Architecture (ZTA) and Post-Quantum Cryptography (PQC) standards, such as ML-KEM (Kyber), to ensure long-term resilience against quantum threats. By synthesizing regulatory trends like the EU Cyber Resilience Act, this research provides a comprehensive framework for transitioning from reactive defense to proactive, intelligent resilience in the IoT ecosystem