Secure IoT Systems in Product Lifecycle Information Management

Existing and upcoming applications of Internet of Things (IoT) show great promise in increasing the level of comfort, efficiency, and automation for human users. These applications require a high-security level to protect users from different types of security threats such as IoT botnets and ransomware. Most of the existing approaches for network security are unable to cope with various limitations of IoT networks, including data heterogeneity and processing power constraints. Although the use of IoT has grown exponentially in recent years, the security of IoT products and users is still often neglected throughout the lifetime of IoT systems. This thesis is one of the first studies that considers IoT security throughout the product lifecycle. Because security is an imperative and ongoing task, it should be started from the earliest stage in the product lifecycle and continued until the final stage. Furthermore, it is vital to ensure the security of not only user clients but also products. However, most current IoT vendors mainly focus on the security requirements of clients, since it is important for them to convince prospective clients that it is safe to adopt their services. For this purpose, the current literature has mostly focused on technologies for safeguarding the security of IoT service clients. Hence, in this thesis, a new security architecture is proposed for IoT that both covers the entire product lifecycle as well as considers product-side and client-side security. By focusing on product-side security, the thesis employs novel machine learning techniques for identifying IoT products in smart environments.