Tuesday, June 2, 2009

Airport Networks: Design, Efficiency and Emergence of Pandemic

The latest pandemic of swine (H1N1) flu has brought forth the issue of the world being a global village and increasing becoming so. A few decades back, when the world was not so well-connected without, particularly, the air-transportation, diseases such as SARS and swine (H1N1) flue would get contained to a relatively smaller geography and die out in short time.

Not the same any more. Before the origin of the disease could be located, much before the scientists are able to find which strain it is of, the human carriers could spread it across the continents, thanks to the dense and fast aviation infrastructure we have developed. A pandemic is thus born. A tool and infrastructure that was meant to serve the mankind is posing a threat of generating
world-wide pandemic owing to human dynamics over the aviation infrastructure.

Is it possible to design airport networks such that while aiming for better efficiency in traffic dynamics, one could simultaneously reduce the chances of possible pandemic emerging over the network?

One paradigm with which to study the aviation infrastructure is that of complex systems, specifically complex networks. It is then possible to model the airport networks using such a model in which the spread of epidemic/pandemic on such a system could be viewed as a flow of information.

In past the aviation infrastructures have been studied at the national
[1-4] as well as world-wide level [5] using complex networks model. There have been studies that model the spread of infectious diseases and predict spread of diseases in future events [6]. In the light of such studies, it would be interesting to know whether the above question can be answered.
A network is said to be assortative if richly connected nodes in it tend to connect to other rich nodes and vice versa. In a disassortatve network the richly connected nodes tend to be connected to poorly connected nodes.
One of the interesting features that comes out from the study of 'efficiency' and 'risks' of airport networks is that there is an apparent dichotomy between the two [7]. Network topology is one aspect that could possibly
be engineered to achieve the desired results. An assortative network is reported to be conducive for information transfer, hence one with weaker resistance to spread of contagious diseases over it. On the other hand such a network is known to be resilient to simple targeted attacks from computational studies [8].

Enhancing airport networks to make them epidemic/pandemic tolerant, by tweaking the topology, would evidently make them prone to targeted attacks. Is there a way out of it? Perhaps there is [7].

The idea is to treat topological assortativity different from that of dynamic (traffic-generated) assortativity. The former is responsible for efficiency of the network (carrying the passengers over the network) and the latter enumerates a measure of human-to-human interaction on the airport network.

Thus a possibly ideal solution would be to tweak the topology to make it assortative while striving to achieve disassortative traffic dynamics profile over the network. Such a network would expectedly have a resilient topology against targeted attacks while at the same time possibly restraining percolation of infectious diseases across it. The details of how
exactly to achieve such a "network state" in a geopoliticaly divided world-wide airport network leads to interesting thoughts to ponder upon.
[1] Ganesh Bagler. Analysis of Airport Network of India as a complex weighted network. Physica A, 387, 2972–2980 (2008).

[2] W. Li and X. Cai. Statistical analysis of Airport Network of China. Phys. Rev. E, 69, 046106 (2003).

[3] M Guida and F. Maria. Topology of the Italian airport network: A scale-free small world network with a fractal structure? Chaos: Solitons and Fractals, 31, 527–536 (2007).

[4] Carlos Pestana Barrosa and Peter U.C. Dieke. Performance evaluation of italian airports: A data envelopment analysis. J. of Air Transport Management, 13, 184–191 (2007).

[5] Alain Barrat et al., The architecture of complex weighted networks. Proc. Natl. Acad. Sci. (USA), 101, 3747–3752 (2004).

[6] Vittoria Colizza et al., The role of the airline transportation network in the prediction and predictability of global epidemics. Proc. Natl. Acad. Sci. (USA), 103, 2015–2020 (2006).

[7] Ganesh Bagler, "Complex network view of performance and risks on airport network", Nova Science Publishers, USA, ISBN: 978-1-60692-1 (2009).

[8] M. E. J. Newman. Assortative mixing in networks. Phys. Rev. Lett., 89, 208701 (2002).

Also See: http://www.industry.siemens.com/Airports/en/
(Siemens is one of the companies that is working on various aspects of airport networks in view of design and growth of "Future of Airports". During their word-wide empirical study, representatives from Siemens interviewed me and discussed various aspects of airport networks, especially Airport Network of India.)

No comments: