Structural Properties of US Flight Network
摘要
We analyse the directed, weighted and evolutionary US flight network, in which vertices are the airports and the flights connecting two airports represent the edges. It is shown that such a network displays two important features recently found in small-world networks. First, the average shortest-path length is 2.4s, the clustering coefficient of the entire network, 0.618, is greatly larger than that of the random networks with the same N (system size) and (k) (average degree), 0.065. We study the detailed flight information both in a week and on a whole. In both the cases, using the Pareto distributions, we find the degree distributions follow two-segment power laws. The weight distributions have power-law tails with the exponents of around 2, and the clustering coefficient distributions follow linear distributions.