The Center for Education and Research in Information Assurance and Security (CERIAS)

The Center for Education and Research in
Information Assurance and Security (CERIAS)

Conflict/Error Prevention, Detection, and Recovery over Networks

Principal Investigator: Shimon Nof

Conflicts and errors are unavoidable in complex networks such as energy grids, supply chains and computer systems. The objective of this research is to design algorithms for effective and automated conflict/error (CE) prevention, detection, and recovery (PDR). The relationships of different CE play an important role: Local CE can propagate to large-scale system damage according to CE dependencies, if not handled correctly and in time. On the other hand, the structural information of the CE network can improve the PDR operations. Constraint-based models are designed based on the complex network theory to define CE dependencies and provide prescriptive abstractions for real-world systems. A centralized algorithm taking advantage of network stracture, a decentralized algorithm enabling parallelism with distributed PDR agents, and hybrid algorithms are designed to prevent, detect, and recover from CEs. The new algorithms use relationships between CE constraints to improve efficiency. Analytical study and simulation experiments on various systems are conducted to validate the new algorithms and compare their performance to that of the traditional algorithm. Results show that for effective PDR, new algorithms shall be used according to several performance measures: Response time, coverage ability, preventability, and damage. The alignment between algorithms and network characteristics, i.e., centralized algorithms for centralized networks and decentralized algorithms for decentralized networks, improves PDR. During the PDR operations, the collaboration between PDR agents also need to be efficiently coordinated and optimized to minimize the cascading effects of CE.

 

Personnel

Other PIs: Xin W. Chen

Students: Hao Zhong

Representative Publications

Keywords: Collaborative control, Complex systems, Detection algorithms, Error detection, Network topologies, Prevention