A large amount of work in regard to the air transportation network analysis has already been done. In the following section we will be presenting a brief overview of some of the interesting published studies that we have encountered during our research. A paper by Dorothy Cheung and Mehmet Gunes dissects the U.S. air transportation network in order to understand its characteristics. In their research, they focus on the basic network analysis by measuring features such as the average path length, degree distribution, clustering, betweeness centrality and resiliency (robustness).
In addition to this, they drew comparisons between their study and earlier air transportation studies. They found that the modern U.S. air infrastructure exhibits same small-world network features similarly to the legacy network, however, when it comes to the power-law distribution, the modern network exhibits only a partial power-law distribution compared to the legacy network which exhibited a complete power-law distribution.
Another study by Massimiliano Zanin and Fabrizio Lillo presents a few different uses of the complex network methods for the characterization of the air transportation infrastructure. In their research, they review a few different papers and compare conclusions in regard to network centrality, degree distributions, and vulnerability. They provide an interesting example of the Ryanair and Lufthansa networks in Europe, with their respective graphs. The conclusion they made was that, at the time of their research, the Ryanair network was more densely connected around the core, while the Lufthansa network was more clustered due to the multiple large hubs affecting its structure. A third research by Verma et al. focuses on identifying the structure of the global flight route data through community detection.
127 our experts are availableright now
They measured the size of closely formed communities based on the flight counts between two destinations (airports). To do this they had ‘used the definition of modularity as introduced by Newman’, which stated that the ‘modularity of a partition is a value that measures the density of links within a community compared to the links that are holding the communities together’. This had ultimately helped them uncover 20 well-connected communities, ‘identifying economically agglomerated regions of the world, such as Middle-East, South-East Asia, Alaska, and Oceania’ . They had also analyzed the connectivity of the network by using two different approaches. In the first approach, they sequentially removed a node (airport) with the highest degree.
This had shown them that upon removing this nodes, the network quickly breaks down into small clusters and the nodes that provide the network with the small-world property also break down. The second approach included the removal of the lowest degree airport which helped them realize that this type of failure causes merely a linear large cluster decay and has no major effects on the rest of the network, as the removed nodes are located on the periphery of the graph. The study conducted by Song and Yeo uses the social network analysis method to conduct an air transportation network analysis of 1,060 airports spread across 173 countries.. They presented two different types of results. The first included the generalized research results and network characteristics for major countries and regions, while the second type of results included the individual network comparison between the United States and China flight route networks.
An article by Lordan and his team focuses on the importance of understanding the structure of the air traffic networks and their vulnerabilities with the goal of assessing potential damages of nonoperational airports and their effect on airlines, countries, and entire continents. In their article, they emphasize the importance of the complex network theory as one of the main means to examine the air traffic network robustness and propose a research agenda with three levels of analysis (global route network, airline alliances, and airlines) that would ultimately ‘help policymakers, airlines, and alliance managers to improve the robustness of the air communications’.
Another article by J. Wang, Mo, F. Wang, and Jin uses a complex network approach to examine the centrality of individual transportation nodes (cities) in China. By measuring the degree, closeness, and betweeness for each one of those nodes, they were able to classify the nodes into three separate groups – nodes directly connected to others, nodes accessible to others, and intermediary nodes.
They concluded that the cumulative degree distribution of this network could be captured by an exponential function, and that the network exhibited some small-world properties such as a small average path length (2.23) and a clustering coefficient of 0.69. They had also found that these metrics were closely correlated with the socio-economic indicators of the cities in which the airports were located, such as passenger volume, population, and and GDP. Further more, they drew a correlation between centrality and its importance on shaping certain socio-economic patterns.
