What makes natural networks great
Join the DZone community and get the full member experience.Join For Free
The world is a hugely connected place, with the networks that form the Internet increasingly efficient and robust. This is due in large part to technologies such as the MPLS network, which reduces network congestion and improves the general user experience.
How do such man made networks compare with the kind of networks we see in nature, such as the neural networks in our brain for instance? A new study has set out to explore just why that is, and what we can learn from nature when designing our own networks.
The paper, which was published in Nature Physics, suggests that a big part of the strength in natural networks is the way that organisms within the network communicate and interact with each other, thus ensuring a rapid and efficient exchange of information. Man made networks on the other hand, can often suffer from cascading failures as a result of relatively small perturbations.
This enhanced state is typically created by the internal structure of natural networks, and the way connections are forged in those networks.
“In this study we have shown that if the interconnections between network nodes are done through nodes with high connectivity within their networks (hubs) and if there is a moderate degree of convergence in the connections between networks, the system of interconnected networks is stable and resistant to failure,” the researchers say.
They suggest that the neural networks found within the brain are setup in such a way that it maximizes the stability of the network. They believe that their findings should make the design of human networks, such as power grids and financial networks, more robust and secure.
“Our results provide not only an answer to the question of why natural networks are more stable than artificial ones, but they also provide a prediction of how structured networks, whether natural or man-made, should be organized in order to acquire stability.”
Opinions expressed by DZone contributors are their own.