Filippo Radicchi, professor of Informatics at the Luddy School of Informatics, Computing, and Engineering, and Minsuk Kim, a Luddy Ph.D. Informatics student, have co-authored a paper that was accepted by the prestigious Physical Review Letters.
Physical Review Letters is the world’s premier physics letter journal and the American Physical Society’s flagship publication. Since 1958, it has helped advance and diffuse physics knowledge by publishing seminal research from Nobel Prize winners and others in all physics-related fields.
In their paper, Shortest-path percolation on random networks, Radicchi and Minsuk propose a bond-percolation model intended to describe consumption, and eventual exhaustion, of resources in transportation networks.
Radicchi and Kim said that from transportation networks to power grid networks, various infrastructure networks serve as the fundamental backbones of our society. Demands are fulfilled by consuming the resources of infrastructure networks with an associated cost.
“One interesting observation is that resource consumption can happen through multiple entities of the network, in other words, through a path,” Radicchi and Kim said. “If any of the entities’ part of the path lacks resources, the demand can be fulfilled via a longer or suboptimal path. Thus, it is crucial to understand the functionality and robustness of various infrastructure networks under path-based resource consumption.”
They said the shortest-path percolation model serves to describe how supplying paths to a population of agents ultimately leads to the depletion of a network’s resources. A key study finding is that allowing agents to have unbounded budgets instead of imposing a cap makes the network more sensitive to external perturbations and leads to a greater waste of its resources.
“Our findings underscore the fundamental importance of relying on path-based rather than standard edge-based models for resource consumption in infrastructure networks,” they said.
The paper was published on July 26.