Decentralised and distributed systems can support business with needs relating to computing and processing capacity in real time. However, these systems are constantly facing challenges from the sophisticated cyber-attack which are evolved and propagated different parts of the system to pose any potential disruption. Additionally, there is communication overhead which makes the implementation of the entire authentication and access control procedure complex and lack of mechanism to tackle the data correlation for privacy breach.  Existing approaches are unlikely to provide an effective access control scheme and effectively detect multi-stage attacks and dynamically re-assess the risks, due to the lack of innovation in capturing and correlating events and associated information and collaborative response.

This project will offer a framework that aims to improve the ability to protect the security and privacy of decentralised and distributed systems by providing a cross domain access control scheme, self-aware system for collaborative intrusion detection, and dynamic risk management taken into consideration of sustainability from the dimension of the consumption of resources and energy. It will facilitate effective collaboration among the different subsystems to prevent widespread disruption due to the cyber-attack and share threat information for creating overall situational awareness. By leveraging blockchain, federated learning, multi-agent architecture, Di4SPDS will develop methods and prototypes for enhancing capability relating to access control, intrusion detection, risk management and collaborative response. 

The Di4SPDS project perfectly aligns with the SPiDDS call, specifically by developing   methods for improving security and privacy of distributed and decentralised systems considering resource and energy usage. Di4SPDS will develop a novel cross-domain blockchain-based authentication and access control protocol that will ensure low computation and communication costs while being resilient against authentication attacks. By leveraging blockchain and semi-asynchronous federated learning, the project will develop a multi-agent and self-aware Intrusion Detection System. Finally, a dynamic risk management and sharing capability that will assess and manage the possible risks with potential impact on individual and multiple systems and share this information with relevant parties.

Project partners

• Lappeenranta-Lahti University of Technology - Finland (coordinator)
• Universidad de Castilla La Mancha - Spain
• Université de Nantes - France
• Firat University - Turkey