Akademik Veri Yönetim Sistemi
JOURNAL OF INFORMATION SECURITY AND APPLICATIONS, cilt.95, 2025 (SCI-Expanded, Scopus)
The rapid growth of the Internet of Things (IoT) has transformed numerous sectors by enabling enhanced connectivity and automation among devices in industrial settings. However, this expansion has brought forward notable security concerns, as Internet enabled and connected devices has become increasingly vulnerable to a variety of cyberattacks. This has elevated the importance of Internet of Things security, necessitating robust defense mechanisms. In this paper, we thoroughly examine Intrusion Detection Systems (IDS) within the context of IoT networks, focusing on the different types of attacks and the corresponding detection methods designed to counteract them. Specifically, we classify IoT-specific threats into categories such as network based, device-level, data-centric, and insider attacks, providing insights into their mechanisms, impacts, and real-world occurrences. To address these threats, various IDS approaches are discussed, including signature based IDS, anomaly based IDS, specification based IDS, and hybrid IDS techniques. We further explore the application of Machine Learning in enhancing IDS capabilities for Internet of Things security. Each method's strengths and limitations are evaluated in terms of accuracy, adaptability, computational efficiency, and scalability. By exploring emerging trends, ongoing challenges, and potential future directions in IDS research for IoT, this study underscores the urgent need for adaptive, scalable, and effective IDS frameworks to protect IoT ecosystems against evolving cyber threats. In addition, this survey provides a critical assessment of the current research landscape, highlighting the fundamental challenges that remain unresolved and outlining future research directions derived both from the existing literature and our own domain-specific analysis.