Privacy and Security in Decentralized Cyber-Physical Systems: A Survey

Abstract

With the growing demand on cutting edge technologies specially in the field of internet of things (IoT) cybersecurity the intersections between blockchain and federated learning (FL) is a promising research field. Blockchain based systems, which are inherently decentralized and trustless, allow local immutable logging for transparent, tamperproof security frameworks that support automated threat response via smart contracts while really being able to understand the integrity of the data being shared across a distributed network of machines. FL represents a powerful new paradigm, allowing organizations to create an accurate intrusion detection system (IDS) without needing to centralize sensitive data, a fundamental problem to solve in healthcare, finance, or Industrial IoT. Both technologies have important synergies within hybrid architectures that offer a balanced alternative by leveraging the privacy of FL and the security and auditability of blockchain to create a viable path to robust, scalable, and reliable cyber defense. There remain numerous challenges yet to be resolved around scalability, interoperability between devices and legacy systems, real-world deployment, and energy efficiency, but the cybersecurity landscape appears to be rapidly evolving to incorporate decentralized technologies. This survey material has highlighted the areas in which progress is being made, outlined the core strengths and limitations of the technologies and approaches employed today, and suggested innovative ideas to pursue in shaping secure, and privacy-conscience, adaptive technologies in forthcoming generation of distributed and networked environments.