Tandon Researcher Explores Network Synchronization and Resiliency

Chaotic may be the perfect adjective for New York subway rush-hour traffic, but dozens of engineers and scientists are currently working within the realm of chaotic, nonlinear systems to better understand how humans, fish, and networks can be controlled or influenced.

NYU Tandon Professor Maurizio Porfiri, director of the Dynamical Systems Laboratory, is one of these scientists exploring how to control the process by which a group can be synchronized to perform certain actions or specific behaviors. Porfiri’s research alongside Russell Jeter and Igor Belykh was recently published in the prestigious journal Chaos and featured as one of the journal’s Editor’s Picks. Their research is funded by the U.S. Army Research Office.

In “Overcoming network resilience to synchronization through non-fast stochastic broadcasting,” the team revealed that a network can be synchronized by a broadcasting oscillator that intermittently transmits signals, thereby overpowering the network’s resiliency to attacks. By examining the interplay between the switching of the broadcasting node and the network dynamics, their study discovered that so-called non-fast switching rates create “windows of opportunity” for synchronized behaviors or “collective dynamics.”

Porfiri explains that this notion of non-fast stochastic broadcasting has implications across natural and technological networks, from schools of fish to a network’s ability to resist control by an external signal.

“Within this paradigm, we imagine a single, isolated unit that is tasked with the goal of taming the dynamics of an entire group. How can it achieve this goal?” Porfiri said. “One possibility is to broadcast information intermittently to the whole group, at a target frequency that is conducive to coordination. In this vein, a robotic fish may be tasked with guiding a group of fish to safety and its approach could be to intermittently release a signal to the group, like a visual cue or food, for the group to follow.”

Porfiri noted that this research has an impact across his other projects, including their exploration of creating models for an epidemic’s spreading and their study of human-computer interactions. “In these cases, it is critical to address the problem of promoting or hampering collective dynamics through the use of selected units in the network.”