How to control complex networks
Recently, an MIT Researcher has discovered an Algorithm to identify how many nodes in a network are necessary to control the entire network. The article explains about “Control Theory” and how for many years this field of study was only applied to feedback systems where inputs and outputs to a system are monitored and variables and parameters that define the system are adjusted accordingly. It goes on to explain how an MIT Researcher has only recently figured out how to apply control theory to analyzing complex self assembling networks such as the internet. An algorithm is constructed that figures out how many nodes need to be targeted in order to control the entire network. Finally the algorithm is studied to understand what feature in a specific network influences how many “driver” nodes are required. This feature is unique to every network. The property he found was known as “degree of distribution” which describes the number of connections per node concluding that sparse networks require many nodes (~90 %) to control the whole network and dense networks require few nodes (~ 10%). In the future they plan to study biological networks and hope to identify vulnerabilities in the network that are target points for antibiotics
This article intrigued me and reminded me of the graph theory we learned in class related to networks. We studied that certain nodes have strong influence on the network and some do not. We also studied the network topologies required to portray these characteristics. The principles of Strong and Weak Ties and Triadic closure show that the network topology of a graph alone can tell us a surprising amount about the characteristics of something like social network or a biological network. These features we studied in class are the fundamental building blocks to learning how to analyze networks on a higher level and using innovative new techniques to break down and identify important data on networks.
Article Link : http://web.mit.edu/newsoffice/2011/network-control-0512.html