A network is a description of the connections that allow interactions and influences between parts of a complex system. It is also used to refer to the parts along with the connections, i.e. the system as a whole, when considering the effects of these connections.

There are various types of networks:

transportation networks—the roads of a city,

communication networks—the telephone network,

utility networks—the electric power grid of a country,

supply networks—the food supply system of a city,

networks of molecular reactions—metabolic networks,

networks of cells—neural networks,

networks of computers—the internet, and

social networks—the people in a company, or in a community.

An important property of the network is its topology: i.e. which elements are directly connected to which elements. Then, more specific properties of the connections can be discussed. Each connection of the network can be characterized by properties such as its strength of influence, capacity, etc.

All networks can be thought of as influence networks: the state of the parts that are connected by the network affect each other through the network. Conceptually, we can distinguish between networks that transfer materials (supply, transportation, utility, chemical), and networks that transfer information (neural, communication, computer, social).

How we think about the network often depends on how complex is the behavior of the network itself. For example, we often think about the transportation system as a system in its own right, even though it is a network that connects the socioeconomic system. In this case, we think about traffic jams and not about the role of roads as connectors. In other cases, we focus on how the parts behave as a result of their connections through the network, or how the system as a whole behaves due to the parts and the network.

Many of the most commonly studied networks assume that they are connecting essentially similar parts, with connections that are in some sense similar. Hence the focus is only on the topology of the network which characterizes its behavior. However, more generally, real networks connect dissimilar parts in dissimilar ways.

As part of the study of complex systems, the general understanding that we gain of how networks behave can be transferred between various kinds of systems, whether they are physical, biological, social or engineered. It is useful to think about behaviors that are common to different kinds of networks and behaviors that are different.

See also:

Optimization of Robustness and Connectivity in Complex Networks
Benjamin Shargel,Hiroki Sayama, Irving R. Epstein and Yaneer Bar-Yam
Phys. Rev. Lett. 90, 068701 (2003)

 

Copyright © 2003 Yaneer Bar-Yam All rights reserved.