Locating the source node of diffusion process in cyber-physical networks via minimum observers.

Area of Science:

• Network Science
• Information Diffusion
• Cybersecurity

Background:

• Accurate source localization is vital for cybersecurity, rumor detection, and disease surveillance.
• Current methods for selecting observers are often random or heuristic, lacking efficiency.
• A method to identify the minimum observer set for precise source localization in cyber-physical networks is needed.

Purpose of the Study:

• To propose a novel greedy optimization algorithm for identifying the minimum set of observers for accurate source localization.
• To address the knowledge gap in efficient observer selection for information diffusion source identification.

Main Methods:

• A greedy optimization algorithm analyzing propagation delay differences.
• Extensive simulations on synthetic and empirical networks.
• Integration with the diffusion-back method for robustness.

Main Results:

• The proposed method significantly decreases the number of observers needed (10%-20% of nodes).
• Conventional random selection requires 2-3 times more observers.
• Periphery nodes with single connections are identified as essential observers.
• The combined approach demonstrates robustness against noise.

Conclusions:

• The greedy optimization algorithm provides an efficient method for minimum observer set selection in source localization.
• This approach offers substantial improvements over random and heuristic observer selection strategies.
• The method enhances the accuracy and efficiency of identifying information diffusion origins across various network types.