Adaptive Security Framework For Iot: Utilizing AI And ML To Counteract Evolving Cyber Threats

Abstract

With the rapid growth of the Internet of Things (IoT), new security challenges continue to emerge over time due to the nature of IoT. It becomes challenging to apply a holistic security approach, and existing security measures may become insufficient in protecting IoT devices upon facing new threat categories or sophisticated attack methods. This chapter proposes an adaptive security framework that focuses on machine learning (ML) and artificial intelligence (AI) methodologies that help to address and counteract the evolving IoT threat categories. Based on case studies, the proposed adaptive security framework demonstrates an obvious improvement in terms of attack detection capability when it comes to using AI and ML classifications. The deployed security mechanism can address and correctly classify new attack patterns without redefining them explicitly in the security system.As the Internet of Things (IoT) is evolving to become the enabler of smart cities and smart businesses, its widespread application brings undeniable socio-economic value. However, the fast growth of IoT is also paired with numerous challenges, especially when discussing the concepts from a security standpoint. The traditional security mechanisms designed for computers, servers, or data centers are not always applicable in the IoT domain — given its heterogeneous device characteristics, severely constrained resources, e.g., limited memory, computation power, and little to no space for security hardware.